Crystal bar state dynamic detection method and device, equipment and storage medium

Abstract

The invention relates to a dynamic detection method, device and equipment for a crystal bar state and a storage medium. The method comprises the following steps: monitoring and collecting pressure drop analog quantity data generated by a crystal bar in a single crystal furnace in different process stages in real time; analyzing and processing the voltage drop analog quantity data, constructing a dynamic model according to an analysis result, and identifying crystal bar state information of the crystal bar in different process stages; acquiring equipment parameter data, inputting the voltage drop analog quantity data and the equipment parameter data into a preset crystal bar state prediction model to obtain ideal state information of the crystal bar, and performing comparative analysis on the ideal state information of the crystal bar and the state information of the crystal bar to generate state difference information; and in combination with the voltage drop analog quantity data and the historical state data which are collected in real time, analyzing the state difference information, and generating a corresponding state analysis report. The method has the effect of improving the accuracy of crystal bar state detection in the production process.

Description

Technical Field

[0001] The present application relates to the technical field of semiconductor manufacturing, and in particular to a method, device, equipment and storage medium for dynamically detecting a crystal rod state. Background Art

[0002] Currently, in the production process of single crystal furnace crystal rods, the crystal rod state detection technology usually relies on a single process parameter for monitoring and analysis, such as relying only on single variable data of temperature or pressure for judgment. However, due to the complex and changeable state of the crystal rod in different process stages (such as welding, seeding, shouldering, and finishing), a single parameter cannot fully reflect the actual state of the crystal rod, resulting in insufficient accuracy of the detection result. Especially in the dynamic process, the pressure drop simulation data of the crystal rod will fluctuate significantly with the change of the process stage, and the existing technology fails to effectively process and analyze these dynamic data, and cannot accurately predict the state of the crystal rod, making it difficult to timely discover potential process problems.

[0003] The above-mentioned prior art solutions have the following defects: In order to improve the accuracy of crystal rod state detection during the production process, there is room for improvement. Summary of the invention

[0004] In order to improve the accuracy of crystal rod state detection during the production process, the present application provides a dynamic detection method, device, equipment and storage medium for the crystal rod state.

[0005] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A method for dynamically detecting a crystal ingot state, the method comprising: Real-time monitoring and collection of simulated pressure drop data generated by the crystal rods in the single crystal furnace at different process stages; Analyzing and processing the voltage drop simulation data, and constructing a dynamic model according to the analysis results to identify the crystal rod state information of the crystal rod at different process stages; Acquire device parameter data, and input the voltage drop simulation data and the device parameter data into a preset crystal rod state prediction model to obtain ideal crystal rod state information, compare and analyze the ideal crystal rod state information with the crystal rod state information, and generate state difference information; The state difference information is analyzed by combining the voltage drop simulation data and the historical state data collected in real time to generate a corresponding state analysis report.

[0006] By adopting the above technical scheme, by real-time monitoring and collecting the pressure drop simulation data generated by the crystal rod in the single crystal furnace at different process stages, and analyzing and processing the data, a dynamic model is constructed to identify the state information of the crystal rod at different process stages, which can comprehensively capture and reflect the actual state of the crystal rod in the production process, thereby improving the accuracy of crystal rod state detection; by inputting the pressure drop simulation data and equipment parameter data into a preset crystal rod state prediction model, the ideal state information of the crystal rod is obtained, and it is compared and analyzed with the actual state to generate state difference information, which can timely discover potential problems in the crystal rod production process, thereby improving the effectiveness of process control and the stability of production quality; combining the real-time collected data with the historical state data for analysis to generate a corresponding state analysis report, which can provide a scientific basis for decision-making in the production process, thereby improving the level of refined management of production.

[0007] In one example, the present application may be further configured as follows: analyzing and processing the voltage drop simulation data, and constructing a dynamic model according to the analysis results, specifically including: Using an adaptive filtering algorithm to smooth the voltage drop analog data to obtain standard analog data; Data mining technology is used to extract key features from the standard analog quantity data, and the dynamic model describing the state of the crystal rod is constructed based on the key features.

[0008] By adopting the above technical solution, the voltage drop analog data is smoothed by using an adaptive filtering algorithm to obtain standard analog data, which can effectively reduce noise interference in the data, thereby improving the accuracy and stability of data processing; data mining technology is used to extract key features from the standard analog data, and a dynamic model describing the crystal rod state is constructed based on these features, which can accurately reflect the state changes of the crystal rod in different process stages, thereby improving the accuracy and comprehensiveness of crystal rod state monitoring.

[0009] In one example, the present application may be further configured as follows: the identifying the crystal ingot status information of the crystal ingot at different process stages specifically includes: Identify the output result of the dynamic model through a pattern recognition algorithm and generate a corresponding state curve; Based on the state curve, the state information of the crystal rod at different process stages is obtained.

[0010] By adopting the above technical solution, the output results of the dynamic model are identified through the pattern recognition algorithm, and the corresponding state curve is generated, which can intuitively reflect the changing trend of the crystal rod state, thereby improving the accuracy of identifying the crystal rod state information at different process stages; the crystal rod state information at different process stages is obtained based on the state curve, which can accurately distinguish the state characteristics of the crystal rod at each process stage, thereby improving the control accuracy of the production process and the quality management level.

[0011] In one example, the present application may be further configured as follows: the construction of the preset crystal rod state prediction model specifically includes: Acquire historical equipment parameter data and historical voltage drop simulation data and historical crystal rod status data corresponding to the historical equipment parameter data, and classify the historical equipment parameter data and the historical voltage drop simulation data according to the historical crystal rod status data to obtain a model training data set; The preset machine learning algorithm is iteratively trained according to the model training data set to obtain the crystal rod state prediction model.

[0012] By adopting the above technical solution, by acquiring historical equipment parameter data, historical voltage drop simulation data and historical crystal rod status data, and classifying the historical equipment parameter data and historical voltage drop simulation data according to the historical crystal rod status data, a model training data set is obtained, which can ensure the comprehensiveness and accuracy of the model training data, thereby improving the prediction ability and applicability of the crystal rod status prediction model; by iteratively training the preset machine learning algorithm, a crystal rod status prediction model is obtained, which can improve the model's prediction accuracy for the crystal rod status by continuously optimizing the model parameters, thereby improving the quality control and fault prevention capabilities in the crystal rod production process.

[0013] In one example, the present application may be further configured as follows: the historical equipment parameter data and the historical voltage drop simulation data are classified according to the historical crystal rod state data to obtain a model training data set, wherein the model training data set includes a positive training data set and a negative training data set, specifically including: Generate a crystal ingot state curve according to the historical crystal ingot state data, and compare the crystal ingot state curve for abnormal points according to a preset crystal ingot state linear change strategy; According to the comparison result, the historical equipment parameter data corresponding to the abnormal state point and the historical voltage drop simulation data corresponding to the historical equipment parameter data are encapsulated as the negative training data set, and other data except the abnormal state point are encapsulated as the positive training data set.

[0014] By adopting the above technical solution, by generating a crystal rod state curve according to historical crystal rod state data, and comparing the crystal rod state curve for abnormal points according to a preset crystal rod state linear change strategy, it is possible to effectively identify abnormal state points in historical data, thereby improving the quality of model training data; by encapsulating the data corresponding to the abnormal state points as a negative training data set and encapsulating the normal data as a positive training data set, it is possible to effectively distinguish between normal states and abnormal states, thereby improving the robustness and prediction accuracy of the crystal rod state prediction model.

[0015] In one example, the present application may be further configured as follows: combining the voltage drop analog data and historical status data collected in real time, analyzing the status difference information and generating a corresponding status analysis report, specifically including: Using a time series analysis method, the real-time collected pressure drop simulation data is compared with the historical state data to identify the law of long-term process changes; Based on the trend comparison result, the process problems caused by the state difference information are obtained, and the state analysis report is generated.

[0016] By adopting the above technical solution and using the time series analysis method to compare the real-time collected voltage drop simulation data with the historical status data, the long-term process change rules in the crystal rod production process can be identified, thereby improving the sensitivity and prediction ability to process state changes; identifying process problems based on the trend comparison results and generating a status analysis report can provide a scientific basis for process optimization and improvement, thereby improving the stability of the production process and the reliability of product quality.

[0017] The second object of the invention is achieved by the following technical solutions: A dynamic detection device for a crystal rod state, the dynamic detection device for a crystal rod state comprising: The data monitoring and collection module is used to monitor and collect the pressure drop simulation data generated by the crystal rods in the single crystal furnace at different process stages in real time; A dynamic model building module is used to analyze and process the voltage drop simulation data, and to build a dynamic model according to the analysis results to identify the crystal rod state information of the crystal rod at different process stages; A state prediction and comparison module is used to obtain device parameter data, and input the voltage drop simulation data and the device parameter data into a preset crystal rod state prediction model to obtain ideal state information of the crystal rod, and compare and analyze the ideal state information of the crystal rod with the crystal rod state information to generate state difference information; The status analysis report generating module is used to analyze the status difference information by combining the voltage drop analog data and the historical status data collected in real time, and generate a corresponding status analysis report.

[0018] By adopting the above technical scheme, by real-time monitoring and collecting the pressure drop simulation data generated by the crystal rod in the single crystal furnace at different process stages, and analyzing and processing the data, a dynamic model is constructed to identify the state information of the crystal rod at different process stages, which can comprehensively capture and reflect the actual state of the crystal rod in the production process, thereby improving the accuracy of crystal rod state detection; by inputting the pressure drop simulation data and equipment parameter data into a preset crystal rod state prediction model, the ideal state information of the crystal rod is obtained, and it is compared and analyzed with the actual state to generate state difference information, which can timely discover potential problems in the crystal rod production process, thereby improving the effectiveness of process control and the stability of production quality; combining the real-time collected data with the historical state data for analysis to generate a corresponding state analysis report, which can provide a scientific basis for decision-making in the production process, thereby improving the level of refined management of production.

[0019] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for dynamically detecting the state of a crystal rod when executing the computer program.

[0020] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for dynamically detecting the state of a crystal rod.

[0021] In summary, this application includes the following beneficial technical effects: 1. By real-time monitoring and collecting the analog pressure drop data generated by the crystal rods in the single crystal furnace at different process stages, and analyzing and processing the data, a dynamic model is constructed to identify the state information of the crystal rods at different process stages, which can comprehensively capture and reflect the actual state of the crystal rods in the production process, thereby improving the accuracy of crystal rod state detection; by inputting the analog pressure drop data and equipment parameter data into the preset crystal rod state prediction model, the ideal state information of the crystal rod is obtained, and it is compared and analyzed with the actual state to generate state difference information, which can timely discover potential problems in the crystal rod production process, thereby improving the effectiveness of process control and the stability of production quality; combining the real-time collected data with the historical state data for analysis, generating a corresponding state analysis report, which can provide a scientific basis for decision-making in the production process, thereby improving the level of refined management of production; 2. By using an adaptive filtering algorithm to smooth the voltage drop analog data to obtain standard analog data, the noise interference in the data can be effectively reduced, thereby improving the accuracy and stability of data processing; using data mining technology to extract key features from the standard analog data, and constructing a dynamic model describing the state of the crystal rod based on these features, it can accurately reflect the state changes of the crystal rod at different process stages, thereby improving the accuracy and comprehensiveness of the crystal rod state monitoring; 3. By acquiring historical equipment parameter data, historical voltage drop simulation data and historical crystal rod status data, and classifying the historical equipment parameter data and historical voltage drop simulation data according to the historical crystal rod status data, a model training data set is obtained, which can ensure the comprehensiveness and accuracy of the model training data, thereby improving the prediction ability and applicability of the crystal rod status prediction model; by iteratively training the preset machine learning algorithm, a crystal rod status prediction model is obtained, which can improve the model's prediction accuracy for the crystal rod status by continuously optimizing the model parameters, thereby improving the quality control and fault prevention capabilities in the crystal rod production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of a method for dynamically detecting a crystal rod state in an embodiment of the present application; Figure 2 is a flowchart for implementing step S20 in the dynamic detection method of the crystal rod state in one embodiment of the present application; Figure 3 is another implementation flow chart of step S20 in the dynamic detection method of the crystal rod state in one embodiment of the present application; Figure 4 It is a flow chart for realizing a crystal rod state prediction model in a dynamic detection method of a crystal rod state in an embodiment of the present application; Figure 5 is a flowchart for implementing step S301 in the method for dynamically detecting the state of a crystal rod in one embodiment of the present application; Figure 6 is a flowchart for implementing step S40 in the method for dynamically detecting the state of a crystal rod in one embodiment of the present application; Figure 7 It is a principle block diagram of a dynamic detection device for a crystal rod state in an embodiment of the present application; Figure 8 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The present application is further described in detail below in conjunction with the accompanying drawings.

[0024] In one embodiment, if Figure 1 As shown, the present application discloses a method for dynamic detection of a crystal rod state, which specifically includes the following steps: S10: Real-time monitoring and collection of pressure drop simulation data generated by the crystal rod in the single crystal furnace at different process stages.

[0025] Specifically, during the operation of the single crystal furnace, the system collects the analog data of the pressure drop of the crystal rod in real time through sensors deployed at different positions, continuously monitors the changes in pressure drop at each process stage, and ensures the integrity and accuracy of the data.

[0026] S20: Analyze and process the voltage drop simulation data, and build a dynamic model according to the analysis results to identify the crystal rod state information at different process stages.

[0027] Specifically, the collected voltage drop analog data are preprocessed, including noise filtering, data smoothing and other steps to improve data quality, and the data are smoothed using an adaptive filtering algorithm to eliminate short-term fluctuations in the data to obtain standardized voltage drop analog data. Data mining technology is used to extract key features from the standardized data, and a dynamic model that can describe the state of the crystal rod is constructed. The state information of the crystal rod in different process stages can be identified from the dynamic model.

[0028] S30: Acquire device parameter data, and input the voltage drop simulation data and the device parameter data into a preset crystal rod state prediction model to obtain ideal crystal rod state information, compare and analyze the ideal crystal rod state information with the crystal rod state information, and generate state difference information.

[0029] Specifically, relevant equipment parameter data, including temperature, pressure, equipment operating time, etc., are obtained from the equipment management module, and these equipment parameter data are input into the preset crystal rod state prediction model together with the real-time collected pressure drop simulation data. The model uses these input data to generate ideal state information of the crystal rod, and then compares the ideal state information with the current actual crystal rod state information to identify state differences and generate state difference information reflecting process deviations.

[0030] S40: Analyze the status difference information by combining the real-time collected voltage drop analog data and the historical status data, and generate a corresponding status analysis report.

[0031] Specifically, the real-time collected pressure drop simulation data is matched and compared with the historical status data, and the time series analysis method is used to identify the rules of long-term process changes. Based on the comparison results, the process problems that may be caused by the status difference information are further identified, and a detailed status analysis report is generated, which includes the possible causes of the process problems, the scope of impact and other information, to provide corresponding decision-making support for operators.

[0032] In one embodiment, if Figure 2As shown, in step S20, the pressure drop simulation data is analyzed and processed, and a dynamic model is constructed according to the analysis results, which specifically includes: S21: Use an adaptive filtering algorithm to smooth the voltage drop analog data to obtain standard analog data.

[0033] Specifically, the collected voltage drop analog data is preliminarily screened to remove obvious errors or noise data, and then the screened data is smoothed through an adaptive filtering algorithm. The algorithm can automatically adjust the filtering parameters according to data fluctuations to minimize data noise and retain key trends, and finally generate smoother and more coherent standard analog data to facilitate subsequent feature extraction and analysis.

[0034] S22: Data mining technology is used to extract key features from standard analog data, and a dynamic model describing the state of the crystal rod is constructed based on the key features.

[0035] Specifically, data mining technology is used to conduct in-depth analysis of the smoothed standard analog data, and key features such as average value, peaks and troughs, slope, etc. are extracted to reflect the state changes of the crystal rod in different process stages. Based on the extracted key features, a dynamic model for describing the state of the crystal rod is constructed. This model can accurately capture subtle changes in the process and provide a basis for subsequent process control and optimization.

[0036] In one embodiment, if Figure 3 As shown, in step S20, the crystal ingot status information at different process stages is identified, which specifically includes: S23: Identify the output result of the dynamic model through a pattern recognition algorithm and generate a corresponding state curve.

[0037] Specifically, the output results generated by the dynamic model are analyzed using a pattern recognition algorithm to identify the state characteristics of the crystal rod at different process stages. The algorithm can automatically distinguish the normal and abnormal states of the crystal rod and generate corresponding state curves to show the state change trend of the crystal rod during the entire process, helping operators to understand and grasp the process progress in a timely manner.

[0038] S24: Based on the state curve, the state information of the crystal rod at different process stages is obtained.

[0039] Specifically, according to the generated state curve, it is divided into segments corresponding to different process stages. By analyzing the curve characteristics of each stage, the specific state information of the crystal rod, such as contact state and lifting and breaking state, is identified, which provides an important reference for evaluating process quality and provides data support for subsequent process optimization.

[0040] In one embodiment, if Figure 4As shown, in step S30, the construction of a preset crystal rod state prediction model specifically includes: S301: Obtain historical equipment parameter data and historical voltage drop analog data and historical crystal rod status data corresponding to the historical equipment parameter data, and classify the historical equipment parameter data and the historical voltage drop analog data according to the historical crystal rod status data to obtain a model training data set.

[0041] Specifically, historical equipment parameter data related to crystal rod production, as well as the corresponding historical voltage drop simulation data and historical crystal rod status data are extracted from the historical database and automatically classified into data sets for different process stages. At the same time, the data sets for different process stages are also classified according to the historical crystal rod status data to generate a data set for machine learning model training, laying the foundation for accurate prediction of subsequent models.

[0042] S302: Iteratively train a preset machine learning algorithm according to a model training data set to obtain a crystal rod state prediction model.

[0043] Specifically, using the prepared model training data set, appropriate machine learning algorithms such as support vector machines and neural networks are selected for iterative training. During the training process, the system continuously adjusts the model parameters to improve the model's prediction accuracy. After multiple rounds of training, when the error value between the output data obtained from the input data and the established result data does not exceed 5%, the final crystal rod state prediction model is obtained. This model can predict the ideal state of the crystal rod based on real-time data.

[0044] In one embodiment, if Figure 5 As shown, in step S301, the historical equipment parameter data and the historical voltage drop simulation data are classified according to the historical crystal rod state data to obtain a model training data set, which includes a positive training data set and a negative training data set, specifically including: S3011: Generate a crystal ingot state curve according to historical crystal ingot state data, and compare the crystal ingot state curve for abnormal points according to a preset crystal ingot state linear change strategy.

[0045] Specifically, state curves reflecting historical crystal rod state changes are generated based on historical crystal rod state data. These curves show the performance of the crystal rod at different process stages. Then, a preset crystal rod state linear change strategy is applied to analyze these curves to identify abnormal points in the curves. Abnormal points usually represent abnormal conditions that occur in the process, that is, the crystal rod does not grow, contact, break, etc. in an ideal state.

[0046] S3012: According to the comparison result, the historical equipment parameter data corresponding to the abnormal state point and the historical voltage drop simulation data corresponding to the historical equipment parameter data are encapsulated as a negative training data set, and the other data except the abnormal state point are encapsulated as a positive training data set.

[0047] Specifically, based on the abnormal point comparison results, the corresponding historical equipment parameter data and historical pressure drop simulation data are marked as negative training data sets for training the model to identify and handle process abnormalities. The remaining normal data are encapsulated as positive training data sets as a reference for the model to identify standard process states, so as to help the model better learn and identify abnormalities in the process and improve the model's prediction accuracy.

[0048] In one embodiment, if Figure 6 As shown, in step S40, the state difference information is analyzed and a corresponding state analysis report is generated by combining the real-time collected voltage drop analog data and the historical state data, which specifically includes: S41: Use time series analysis methods to compare the real-time collected pressure drop simulation data with historical status data to identify the patterns of long-term process changes.

[0049] Specifically, the time series analysis method is used to perform trend analysis and comparison between the real-time collected pressure drop simulation data and the historical status data. By identifying the long-term trends in the data, regular changes in the process are discovered, thereby identifying potential problems in the process and providing a basis for process improvement.

[0050] S42: Based on the trend comparison result, the process problems caused by the state difference information are obtained, and a state analysis report is generated.

[0051] Specifically, based on the trend comparison results, specific problems in the process that may cause status differences are identified. These problems usually manifest as abnormal fluctuations in process parameters or gradual degradation of equipment performance. The system records these identified process problems in a status analysis report, which lists in detail the manifestations of the problems, possible causes and other information, providing data support for operators' decision-making.

[0052] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0053] In one embodiment, a dynamic detection device for a crystal rod state is provided, and the dynamic detection device for a crystal rod state corresponds one-to-one to the dynamic detection method for a crystal rod state in the above embodiment. Figure 7As shown, the dynamic detection device for the crystal rod state includes a data monitoring and collection module, a dynamic model building module, a state prediction and comparison module, and a state analysis report generation module. The detailed description of each functional module is as follows: The data monitoring and collection module is used to monitor and collect the pressure drop simulation data generated by the crystal rods in the single crystal furnace at different process stages in real time; A dynamic model building module is used to analyze and process the voltage drop simulation data, and to build a dynamic model based on the analysis results to identify the crystal ingot state information at different process stages; The state prediction and comparison module is used to obtain equipment parameter data, and input the voltage drop simulation data and the equipment parameter data into a preset crystal rod state prediction model to obtain the ideal state information of the crystal rod, compare and analyze the ideal state information of the crystal rod with the crystal rod state information, and generate state difference information; The status analysis report generation module is used to combine the real-time collected voltage drop analog data with the historical status data, analyze the status difference information, and generate a corresponding status analysis report.

[0054] Optionally, the dynamic model building module specifically includes: The data smoothing submodule is used to smooth the voltage drop analog data using an adaptive filtering algorithm to obtain standard analog data; The key feature extraction submodule is used to extract key features from standard analog data using data mining technology, and to build a dynamic model describing the state of the crystal rod based on the key features.

[0055] Optionally, the dynamic model building module specifically includes: The state curve generation submodule is used to identify the output results of the dynamic model through a pattern recognition algorithm and generate a corresponding state curve; The state information acquisition submodule is used to obtain the state information of the crystal rod at different process stages based on the state curve.

[0056] Optionally, the construction of the preset crystal ingot state prediction model specifically includes: A model data classification module is used to obtain historical equipment parameter data and historical voltage drop analog data and historical crystal rod status data corresponding to the historical equipment parameter data, and classify the historical equipment parameter data and the historical voltage drop analog data according to the historical crystal rod status data to obtain a model training data set; The machine learning training module is used to iteratively train the preset machine learning algorithm according to the model training data set to obtain a crystal rod state prediction model.

[0057] Optionally, the model data classification module specifically includes: The abnormal point comparison submodule is used to generate a crystal ingot state curve according to historical crystal ingot state data, and to compare the abnormal points of the crystal ingot state curve according to a preset crystal ingot state linear change strategy; The training data set encapsulation submodule is used to encapsulate the historical equipment parameter data corresponding to the abnormal state point and the historical voltage drop simulation data corresponding to the historical equipment parameter data into a negative training data set according to the comparison results, and to encapsulate other data except the abnormal state point into a positive training data set.

[0058] Optionally, the status analysis report generation module specifically includes: The trend comparison submodule is used to compare the real-time collected pressure drop simulation data with the historical status data using the time series analysis method to identify the law of long-term process changes; The process problem identification submodule is used to obtain process problems arising from status difference information based on trend comparison results and generate a status analysis report.

[0059] For the specific definition of the dynamic detection device of the crystal rod state, please refer to the definition of the dynamic detection method of the crystal rod state in the above text, which will not be repeated here. Each module in the above-mentioned dynamic detection device of the crystal rod state can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0060] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a dynamic detection method for a crystal rod state is implemented.

[0061] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: Real-time monitoring and collection of simulated pressure drop data generated by the crystal rods in the single crystal furnace at different process stages; Analyze and process the voltage drop simulation data, and build a dynamic model based on the analysis results to identify the crystal ingot status information at different process stages; Acquire equipment parameter data, and input the voltage drop simulation data and the equipment parameter data into a preset crystal rod state prediction model to obtain ideal crystal rod state information, compare and analyze the ideal crystal rod state information with the crystal rod state information, and generate state difference information; Combine the real-time collected voltage drop simulation data with the historical status data, analyze the status difference information, and generate a corresponding status analysis report.

[0062] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: Real-time monitoring and collection of simulated pressure drop data generated by the crystal rods in the single crystal furnace at different process stages; Analyze and process the voltage drop simulation data, and build a dynamic model based on the analysis results to identify the crystal ingot status information at different process stages; Acquire equipment parameter data, and input the voltage drop simulation data and the equipment parameter data into a preset crystal rod state prediction model to obtain ideal crystal rod state information, compare and analyze the ideal crystal rod state information with the crystal rod state information, and generate state difference information; Combine the real-time collected voltage drop simulation data with the historical status data, analyze the status difference information, and generate a corresponding status analysis report.

[0063] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0064] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0065] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for dynamic detection of crystal ingot state, characterized in that: The dynamic detection method of the crystal rod state includes: Real-time monitoring and collection of simulated pressure drop data generated by the crystal rods in the single crystal furnace at different process stages; Analyzing and processing the voltage drop simulation data, and constructing a dynamic model according to the analysis results to identify the crystal rod state information of the crystal rod at different process stages; Acquire device parameter data, input the voltage drop simulation data and the device parameter data into a preset crystal rod state prediction model to obtain ideal crystal rod state information, compare and analyze the ideal crystal rod state information with the crystal rod state information, and generate state difference information; The state difference information is analyzed by combining the voltage drop simulation data and the historical state data collected in real time to generate a corresponding state analysis report.

2. The method for dynamic detection of the crystal rod state according to claim 1, characterized in that: The analyzing and processing of the pressure drop simulation data and constructing a dynamic model according to the analysis results specifically include: Using an adaptive filtering algorithm to smooth the voltage drop analog data to obtain standard analog data; Data mining technology is used to extract key features from the standard analog quantity data, and the dynamic model describing the state of the crystal rod is constructed based on the key features.

3. The method for dynamic detection of crystal ingot state according to claim 1, characterized in that: The identifying the crystal ingot status information of the crystal ingot at different process stages specifically includes: Identify the output result of the dynamic model through a pattern recognition algorithm and generate a corresponding state curve; Based on the state curve, the state information of the crystal rod at different process stages is obtained.

4. The method for dynamic detection of crystal ingot state according to claim 1, characterized in that: The construction of the preset crystal rod state prediction model specifically includes: Acquire historical equipment parameter data and historical voltage drop simulation data and historical crystal rod status data corresponding to the historical equipment parameter data, and classify the historical equipment parameter data and the historical voltage drop simulation data according to the historical crystal rod status data to obtain a model training data set; The preset machine learning algorithm is iteratively trained according to the model training data set to obtain the crystal rod state prediction model.

5. The method for dynamic detection of the crystal ingot state according to claim 4, characterized in that: The historical equipment parameter data and the historical voltage drop simulation data are classified according to the historical crystal rod state data to obtain a model training data set, wherein the model training data set includes a positive training data set and a negative training data set, specifically including: Generate a crystal ingot state curve according to the historical crystal ingot state data, and compare the crystal ingot state curve for abnormal points according to a preset crystal ingot state linear change strategy; According to the comparison result, the historical equipment parameter data corresponding to the abnormal state point and the historical voltage drop simulation data corresponding to the historical equipment parameter data are encapsulated as the negative training data set, and other data except the abnormal state point are encapsulated as the positive training data set.

6. The method for dynamic detection of crystal ingot state according to claim 1, characterized in that: The combining of the voltage drop simulation data and the historical status data collected in real time, analyzing the status difference information and generating a corresponding status analysis report specifically includes: Using a time series analysis method, the real-time collected pressure drop simulation data is compared with the historical state data to identify the law of long-term process changes; Based on the trend comparison result, the process problems caused by the state difference information are obtained, and the state analysis report is generated.

7. A dynamic detection device for a crystal rod state, characterized in that: The dynamic detection device for the crystal rod state comprises: The data monitoring and collection module is used to monitor and collect the pressure drop simulation data generated by the crystal rods in the single crystal furnace at different process stages in real time; A dynamic model building module is used to analyze and process the voltage drop simulation data, and to build a dynamic model according to the analysis results to identify the crystal rod state information of the crystal rod at different process stages; A state prediction and comparison module is used to obtain device parameter data, and input the voltage drop simulation data and the device parameter data into a preset crystal rod state prediction model to obtain ideal state information of the crystal rod, and compare and analyze the ideal state information of the crystal rod with the crystal rod state information to generate state difference information; The status analysis report generating module is used to analyze the status difference information by combining the voltage drop analog data and the historical status data collected in real time, and generate a corresponding status analysis report.

8. The dynamic detection device for the crystal rod state according to claim 7, characterized in that: The dynamic model building module specifically includes: A data smoothing processing submodule is used to use an adaptive filtering algorithm to smooth the voltage drop analog data to obtain standard analog data; The key feature extraction submodule is used to extract key features from the standard analog data using data mining technology, and to construct the dynamic model describing the state of the crystal rod according to the key features.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for dynamically detecting the state of the crystal rod as claimed in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method for dynamically detecting the state of a crystal rod as claimed in any one of claims 1 to 6 are implemented.

Images