Fault prediction method, device and equipment for turbo molecular pump and medium

Abstract

The invention relates to the technical field of molecular pumps, and discloses a fault prediction method, device, equipment and medium for a turbo molecular pump, the method comprises the following steps: collecting operation data of the turbo molecular pump according to a set sampling frequency, and transmitting the operation data through a data transmission network; performing data cleaning, denoising, normalization and format conversion processing on the historical operation data; the health index value of the turbo molecular pump is calculated according to the processed historical operation data, whether the turbo molecular pump breaks down or not is judged according to the health index value, and historical fault data are obtained; according to the method, the fault prediction model is established based on the historical fault data and the real-time operation data, the fault state of the turbo molecular pump is predicted through the fault prediction model, the fault prediction model is established to predict possible faults of the turbo molecular pump, and therefore potential fault hidden dangers can be found timely and accurately, and the fault state of the turbo molecular pump can be predicted. And a more reliable basis is provided for the maintenance decision of the turbo molecular pump.

Description

Technical Field

[0001] The present invention relates to the technical field of molecular pumps, and in particular to a fault prediction method, device, equipment and medium for a turbomolecular pump. Background Art

[0002] Turbomolecular pumps are widely used in semiconductor manufacturing, vacuum coating, scientific research and other fields. They transfer momentum to gas molecules through high-speed rotating rotor blades to achieve the pumping function. With the continuous advancement of science and technology, the performance, reliability and intelligent control requirements of turbomolecular pumps are getting higher and higher. Therefore, it is very important to predict possible faults of turbomolecular pumps in advance and take effective maintenance measures in time for the maintenance of turbomolecular pumps.

[0003] Traditional turbomolecular pump fault judgment often relies on fixed parameter settings and experience judgment, which is difficult to adapt to different working conditions and load changes, resulting in unstable pumping efficiency. In addition, the current monitoring and early warning of the turbomolecular pump's operating status are limited to the current status, and potential fault hazards cannot be discovered in time, resulting in high maintenance costs after a fault occurs and affecting production progress. Summary of the invention

[0004] In view of this, the present invention provides a fault prediction method, device, equipment and medium for a turbomolecular pump to solve the problem in the prior art that potential fault hazards cannot be discovered in time.

[0005] In a first aspect, the present invention provides a fault prediction method for a turbomolecular pump, the method comprising:

[0006] Collecting the operating data of the turbomolecular pump according to the set sampling frequency, and transmitting the operating data through the data transmission network, the operating data including historical operating data and real-time operating data;

[0007] Perform data cleaning, denoising, normalization and format conversion on historical operation data;

[0008] Calculate the health index value of the turbomolecular pump according to the processed historical operation data, and determine whether the turbomolecular pump fails according to the health index value to obtain historical fault data;

[0009] A fault prediction model is established based on historical fault data and real-time operation data, and the fault prediction model is used to predict the fault state of the turbomolecular pump.

[0010] The present invention collects the operating data of the turbomolecular pump and pre-processes the operating data to evaluate the health status of the turbomolecular pump according to the pre-processed operating data, and predicts the possible faults of the turbomolecular pump through the established fault prediction model, thereby timely and accurately discovering potential fault hazards and providing a more reliable basis for maintenance decisions of the turbomolecular pump.

[0011] In an optional embodiment, after collecting the operating data of the turbomolecular pump according to the set sampling frequency, the method further includes:

[0012] Time stamp the run data.

[0013] The present invention marks the operation data with a timestamp to facilitate tracking of the data in the subsequent analysis process and ensure the integrity and accuracy of the data.

[0014] In an optional implementation, determining whether a turbomolecular pump fails according to a health indicator value includes:

[0015] Set the threshold range of each health indicator of the turbomolecular pump;

[0016] According to the threshold range of each health indicator, a digital twin model of the turbomolecular pump is established using 3D modeling technology and simulation technology. The health indicators include temperature health indicator, pressure health indicator, vibration health indicator and performance efficiency indicator.

[0017] Use digital twin models to calculate health indicators of turbomolecular pumps;

[0018] If the health indicator value exceeds the threshold range of the corresponding health indicator, it is determined that the turbomolecular pump fails.

[0019] The present invention establishes a digital twin model by setting the threshold range of each health indicator, so as to monitor and analyze the operating status of the turbomolecular pump and dynamically evaluate the health status of the turbomolecular pump, so as to judge the fault condition of the turbomolecular pump according to the health indicator value, thereby improving the accuracy of fault diagnosis.

[0020] In an optional implementation, a fault prediction model is established based on historical fault data and real-time operation data, including:

[0021] Use machine learning algorithms and data analysis techniques to establish a fault prediction model for turbomolecular pumps;

[0022] Divide historical fault data and real-time operation data into training data sets and test data sets according to a preset ratio;

[0023] Using the training data set to train the fault prediction model;

[0024] The fault prediction model is validated using the test dataset.

[0025] The present invention establishes a fault prediction model by utilizing machine learning algorithms and data analysis techniques, trains the fault prediction model to continuously optimize the parameters and performance of the fault prediction model, and verifies the fault prediction model to improve the prediction accuracy of the fault prediction model.

[0026] In an optional implementation, using a test data set to verify the fault prediction model includes:

[0027] The generalization ability and effectiveness of the fault prediction model are verified by cross-validation method.

[0028] The present invention verifies the fault prediction model through a cross-validation method to verify the generalization ability and effectiveness of the model, thereby improving the accuracy and reliability of the model prediction.

[0029] In an optional implementation, after predicting the fault state of the turbomolecular pump using the fault prediction model, the method further includes:

[0030] When the fault state prediction result of the turbomolecular pump is a fault, an early warning signal is issued, and the early warning signal includes the fault type, occurrence time and possibility.

[0031] The present invention sends out an early warning signal when the fault state prediction result of the turbomolecular pump is a fault, so as to formulate a maintenance plan for the turbomolecular pump in time, thereby ensuring the normal operation of the turbomolecular pump and reducing maintenance costs and equipment downtime.

[0032] In a second aspect, the present invention provides a fault prediction device for a turbomolecular pump, the device comprising:

[0033] The acquisition module is used to collect the operating data of the turbomolecular pump according to the set sampling frequency, and transmit the operating data through the data transmission network, the operating data including historical operating data and real-time operating data;

[0034] Data processing module, used for data cleaning, denoising, normalization and format conversion of historical operation data;

[0035] A judgment module is used to calculate the health index value of the turbomolecular pump according to the processed historical operation data, and judge whether the turbomolecular pump fails according to the health index value to obtain historical fault data;

[0036] The prediction module is used to establish a fault prediction model based on historical fault data and real-time operation data, and use the fault prediction model to predict the fault state of the turbomolecular pump.

[0037] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the fault prediction method for a turbomolecular pump according to the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0038] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the fault prediction method for a turbomolecular pump according to the first aspect or any corresponding embodiment thereof.

[0039] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions for causing a computer to execute the fault prediction method for a turbomolecular pump according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 is a flow chart of a fault prediction method for a turbomolecular pump according to an embodiment of the present invention;

[0042] Figure 2 is a schematic diagram of a physical turbomolecular pump data acquisition process according to an embodiment of the present invention;

[0043] Figure 3 is a structural schematic diagram of a turbomolecular pump digital twin system according to an embodiment of the present invention;

[0044] Figure 4 is a structural block diagram of a fault prediction device for a turbomolecular pump according to an embodiment of the present invention;

[0045] Figure 5 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0047] According to an embodiment of the present invention, an embodiment of a fault prediction method for a turbomolecular pump is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0048] In this embodiment, a fault prediction method for a turbomolecular pump is provided. Figure 1 is a flow chart of a fault prediction method for a turbomolecular pump according to an embodiment of the present invention, such as Figure 1 As shown, the process includes the following steps:

[0049] Step S101 : collecting operating data of the turbomolecular pump according to a set sampling frequency, and transmitting the operating data through a data transmission network.

[0050] In an embodiment of the present invention, a sensor is installed on each turbomolecular pump to collect various data during the operation of the turbomolecular pump, such as temperature, pressure, vibration, and rotation speed. The sensor has the characteristics of high precision, high reliability, and real-time performance to accurately obtain the operating status of the turbomolecular pump.

[0051] Specifically, temperature sensors are used to monitor the temperature of key components of turbomolecular pumps, such as pump bodies, rotors, bearings, etc. Excessive temperatures may affect the performance and life of the pump, or even cause failure. Temperature sensors can detect temperature anomalies in a timely manner and take appropriate measures for protection and control.

[0052] The turbomolecular pump operation data is collected according to the set sampling frequency. The operation data includes historical operation data and real-time operation data. The collection time should cover different operation stages and working conditions of the pump to ensure the diversity and representativeness of the data. The collected data is stored in the database to establish a historical data archive. The data is regularly sorted and backed up to ensure the security and integrity of the data.

[0053] Build a data transmission network to transmit the collected operation data to the data processing center in real time through the data transmission network. Wireless transmission can be used to ensure the stability and security of data transmission. Among them, data encryption and verification technology is used to ensure the security and reliability of data during transmission and prevent data loss and tampering.

[0054] Step S102, performing data cleaning, denoising, normalization and format conversion on the historical operation data.

[0055] In an embodiment of the present invention, the raw data transmitted to the data processing center is preprocessed, including data cleaning, denoising, normalization and other operations, to remove outliers and noise interference, perform format conversion, and convert the data into a format suitable for analysis and modeling to improve data quality and availability.

[0056] Step S103, calculating the health index value of the turbomolecular pump according to the processed historical operation data, and judging whether the turbomolecular pump fails according to the health index value, to obtain historical fault data.

[0057] In the embodiment of the present invention, a health index system is formulated in combination with the working principle and operation characteristics of the turbomolecular pump. The health index value of the turbomolecular pump is calculated based on the pre-processed historical operation data, and the calculated health index value of the turbomolecular pump is compared with the set health index threshold to determine whether the turbomolecular pump fails.

[0058] Step S104: establishing a fault prediction model based on the historical fault data and the real-time operation data, and using the fault prediction model to predict the fault state of the turbomolecular pump.

[0059] In an embodiment of the present invention, a fault prediction model of a turbomolecular pump is established based on historical fault data and real-time operation data, and the fault prediction model is trained to predict the future fault state of the turbomolecular pump, so that a maintenance plan can be formulated in time according to the prediction results.

[0060] The fault prediction method for a turbomolecular pump provided in this embodiment collects operating data of the turbomolecular pump and preprocesses the operating data to evaluate the health status of the turbomolecular pump based on the preprocessed operating data, and predicts possible faults of the turbomolecular pump through an established fault prediction model, thereby timely and accurately discovering potential fault hazards and providing a more reliable basis for maintenance decisions of the turbomolecular pump.

[0061] In this embodiment, a fault prediction method for a turbomolecular pump is provided, and the process includes the following steps:

[0062] Step S201 : collecting operating data of the turbomolecular pump according to a set sampling frequency, and transmitting the operating data through a data transmission network.

[0063] For details, please see Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0064] Step S202: timestamp the operation data.

[0065] In the embodiment of the present invention, in order to ensure the integrity and accuracy of the data, the collected operation data is timestamped. Specifically, the sensor installed on the turbomolecular pump supports a timestamp marking function, and the collected operation data can be timestamped, or the collected operation data and the timestamp can be stored together. When the operation data is saved, a timestamp field is added to facilitate subsequent analysis and processing.

[0066] By time-stamping the running data, it is easy to track the data during subsequent analysis to ensure the integrity and accuracy of the data.

[0067] Step S203, performing data cleaning, denoising, normalization and format conversion processing on the historical operation data.

[0068] For details, please see Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0069] Step S204, calculating the health index value of the turbomolecular pump according to the processed historical operation data, and determining whether the turbomolecular pump fails according to the health index value to obtain historical fault data.

[0070] Specifically, judging whether the turbomolecular pump fails according to the health index value in the above step S204 includes:

[0071] Step S2041, setting the threshold range of each health indicator of the turbomolecular pump.

[0072] Step S2042: Based on the threshold range of each health indicator, a digital twin model of the turbomolecular pump is established using three-dimensional modeling technology and simulation technology.

[0073] Step S2043, using the digital twin model to calculate the health index value of the turbomolecular pump.

[0074] Step S2044: If the health indicator value exceeds the threshold range of the corresponding health indicator, it is determined that the turbomolecular pump fails.

[0075] In the embodiment of the present invention, a health index system is formulated in combination with the working principle and operation characteristics of the turbomolecular pump, including but not limited to temperature health index, pressure health index, vibration health index and performance efficiency index. Each health index should be set with a corresponding threshold range to judge the health status of the turbomolecular pump.

[0076] According to the threshold range of each health indicator, a virtual digital twin model of the turbomolecular pump is created using 3D modeling software and simulation platform. This model can accurately reflect the appearance and internal structure of the physical pump, as well as the movement relationship between its components.

[0077] The collected operation data is associated with the virtual model to achieve real-time synchronization and interaction between the virtual model and the physical entity. The operation status and performance parameters of the physical pump can be obtained through the operation and analysis of the virtual model.

[0078] Specifically, ANSYS software is used to conduct detailed analysis and modeling of the geometric structure, material properties, working principle, etc. of multiple turbomolecular pumps. The shapes, sizes, and relationships of the rotors, stators, blades, and other components of the turbomolecular pumps, as well as the effects of the physical properties of different materials on the pump performance, are considered.

[0079] The digital twin model has the functions of status monitoring and fault diagnosis. The status monitoring and fault diagnosis calculate the health index value of each turbomolecular pump in real time based on the preprocessed data, compare the calculated health index value with the set health index threshold range, and analyze whether the operating status of the turbomolecular pump is normal. If the health index value exceeds the corresponding threshold range, it is determined that the turbomolecular pump has a fault.

[0080] Specifically, the digital twin model also has a health status visualization function, which can display the health status of multiple turbomolecular pumps to users in an intuitive way through a visualization interface. Specifically, the real-time change trend of various health indicators of turbomolecular pumps, historical data records, and comparisons with thresholds can be displayed in the form of charts, images, etc. Users can quickly understand the health status of each pump through the visualization interface and take corresponding measures in a timely manner.

[0081] By setting the threshold range of each health indicator, a digital twin model is established to monitor and analyze the operating status of the turbomolecular pump and dynamically evaluate the health status of the turbomolecular pump. The fault condition of the turbomolecular pump can be judged according to the health indicator value, thereby improving the accuracy of fault diagnosis.

[0082] Step S205: establishing a fault prediction model based on the historical fault data and the real-time operation data, and using the fault prediction model to predict the fault state of the turbomolecular pump.

[0083] Specifically, in the above step S205, a fault prediction model is established based on the historical fault data and the real-time operation data, and the fault prediction model is used to predict the fault state of the turbomolecular pump, including:

[0084] Step S2051, using machine learning algorithms and data analysis techniques to establish a fault prediction model for the turbomolecular pump.

[0085] Step S2052: divide the historical fault data and the real-time operation data into a training data set and a test data set according to a preset ratio.

[0086] Step S2053: train the fault prediction model using the training data set.

[0087] Step S2054: Use the test data set to verify the fault prediction model.

[0088] In an embodiment of the present invention, based on historical fault data and real-time operation data, a fault prediction model for a turbomolecular pump is established using machine learning algorithms and data analysis techniques. Specifically, algorithms such as neural networks, support vector machines, and decision trees can be used to identify and predict the fault modes of turbomolecular pumps.

[0089] The historical fault data and real-time operation data are divided into training data sets and test data sets according to a preset ratio. The fault prediction model is trained using the training data set to continuously optimize the parameters and performance of the fault prediction model to improve the accuracy and reliability of fault prediction. The prediction results of the fault prediction model are then verified using the test data set.

[0090] By using machine learning algorithms and data analysis technology to establish a fault prediction model, the fault prediction model is trained to continuously optimize the parameters and performance of the fault prediction model, and the fault prediction model is verified to improve the prediction accuracy of the fault prediction model.

[0091] Specifically, the verification of the fault prediction model using the test data set in the above step S2051 includes:

[0092] Step S20511, verifying the generalization ability and effectiveness of the fault prediction model through cross-validation method.

[0093] In the embodiment of the present invention, the generalization ability, effectiveness and accuracy of the fault prediction model are verified by using methods such as cross-validation and error analysis. Specifically, the cross-validation can use K-fold cross-validation, dividing the data set into K subsets, using K-1 subsets as training sets each time, and the remaining 1 subset as a validation set, repeating K times, and each subset is used as a validation set for validation.

[0094] According to the verification results, adjust the parameters and algorithms of the fault prediction model, continuously improve the model's training algorithm and strategy, optimize the performance of the fault prediction model, and improve the model's fit and generalization ability to ensure that the model can accurately identify and predict different types of faults.

[0095] The fault prediction model is validated by cross-validation method to verify the generalization ability and effectiveness of the model, thereby improving the accuracy and reliability of the model prediction.

[0096] Step S206: When the fault state prediction result of the turbomolecular pump is a fault, a warning signal is issued.

[0097] In the embodiment of the present invention, the early warning signal includes the fault type, occurrence time and possibility. The fault state of the turbomolecular pump is predicted by using the established fault prediction model. When a possible fault is predicted, an early warning signal is issued in advance to indicate the possible hidden fault. The early warning signal provides information such as the fault type, occurrence time and possibility. The user can formulate a maintenance plan in time according to the early warning signal, arrange maintenance personnel, and prepare the required spare parts.

[0098] According to the health status assessment results and fault prediction information of the turbomolecular pump, combined with factors such as maintenance cost and equipment downtime, potential fault hazards can be discovered in advance and reasonable maintenance decisions can be made. Condition-Based Maintenance (CBM) strategy can be adopted to determine the maintenance time and content according to the actual operating status of the turbomolecular pump to achieve precise maintenance.

[0099] When making maintenance decisions, the correlation and synergy between multiple turbomolecular pumps should also be considered. For pumps on some key production lines, their operational reliability should be prioritized and corresponding emergency maintenance plans should be formulated. At the same time, maintenance resources should be arranged reasonably to avoid conflicts and excessive concentration of maintenance tasks.

[0100] The fault prediction method for a turbomolecular pump provided in this embodiment sends out a warning signal when the fault state prediction result of the turbomolecular pump is a fault, so as to facilitate timely formulation of a maintenance plan for the turbomolecular pump, thereby ensuring the normal operation of the turbomolecular pump and reducing maintenance costs and equipment downtime.

[0101] As a specific application example of the embodiment of the present invention, the following steps are included:

[0102] (I) Implementation environment construction

[0103] 1. Hardware configuration

[0104] like Figure 2Install the required sensors on each turbomolecular pump as shown:

[0105] (1) Temperature sensor: monitors the temperature of key components of the turbomolecular pump, such as the pump body, rotor, bearings, etc. Excessive temperature may affect the performance and life of the pump, or even cause failure. The temperature sensor can detect temperature abnormalities in time and take appropriate measures for protection and control.

[0106] (2) Pressure sensor: used to measure the gas pressure at the inlet and outlet of the turbomolecular pump to monitor the pumping performance and working status of the pump. By collecting pressure data in real time, it can be determined whether the pump is operating normally and whether the required vacuum degree has been achieved.

[0107] (3) Speed ​​sensor: measures the speed of the turbomolecular pump rotor, which is very important for controlling the pumping rate and performance of the pump. By monitoring the speed, it can ensure that the pump operates within the normal operating range and avoid damage to the pump caused by overspeed or underspeed operation.

[0108] (4) Vibration sensor: Detects the vibration of the turbomolecular pump to determine whether the pump is unbalanced, loose, or has other mechanical faults. Excessive vibration may affect the stability and life of the pump, and even cause safety problems. Vibration sensors can detect vibration abnormalities in a timely manner and perform fault diagnosis and maintenance.

[0109] (5) Current sensor: monitors the driving motor current of the turbomolecular pump to understand the load and operating status of the pump. The change in current can reflect the pump's workload, motor failure and other problems. The current sensor can achieve intelligent control and protection of the pump.

[0110] (6) Gas flow sensor: measures the gas flow rate during the turbomolecular pump’s pumping process to evaluate the pump’s pumping efficiency and performance. The gas flow sensor can be used to monitor the pump’s operating status in real time, optimize the pumping process, and improve production efficiency.

[0111] Build a data transmission network: The controller is responsible for receiving sensor data and controlling the operation of the turbomolecular pump, and transmitting the data to the wireless transmission device. The wireless transmission device uses WIFI technology to ensure that the collected data can be transmitted to the Alibaba Cloud service platform in a timely and accurate manner. During the data transmission process, encryption algorithms and data verification mechanisms are used to ensure data integrity and confidentiality, and prevent data loss and tampering during transmission. At the same time, multiple molecular pumps in different areas are connected through wireless transmission equipment.

[0112] 2. Software system development

[0113] Configure Alibaba Cloud service center servers and storage devices for data processing: Alibaba Cloud servers have sufficient computing power and memory capacity to process large amounts of real-time data and run complex algorithm models. Storage devices should be able to reliably store historical data and model files to facilitate subsequent data analysis and query.

[0114] like Figure 3 As shown in the figure, a virtual turbomolecular pump digital twin model platform is constructed, including 3D modeling software, simulation software, and data analysis tools. These tools are used to create a virtual digital model of the turbomolecular pump and to associate and interact with the model and real-time data.

[0115] Develop a visualization platform: To meet user needs, the digital twin data of the turbomolecular pump should be packaged to provide services such as real-time data monitoring, data review, data query, data analysis, data management, data model selection and construction, data processing algorithm selection and construction, etc., to achieve comprehensive data visualization in digital and graphical form. It should include functional modules such as health indicator calculation module, fault prediction model training and prediction module, and maintenance decision-making module.

[0116] At the same time, you can set different pages to meet different user needs:

[0117] (1) For production activity managers, we can provide production process simulation services, equipment resource allocation optimization services, production plan optimization services, and turbomolecular pump scheduling optimization services to improve production efficiency and reduce operating power consumption;

[0118] (2) For professional and technical personnel of turbomolecular pumps, we can provide operation quality prediction services, equipment operation performance evaluation services, turbomolecular pump control optimization services, key component life prediction services, and fault diagnosis services to improve operation performance, process level and equipment quality;

[0119] (3) Immersive remote visualization services and remote control services can be provided to all users to solve the problems of traditional video surveillance methods, such as large data volume, delay, and inability to observe specific locations, stimulate design inspiration, and improve management and control efficiency.

[0120] 2. Data collection and model training

[0121] 1. Data collection and accumulation

[0122] During the normal operation of the turbomolecular pump, the data acquisition system collects real-time operating data such as temperature, pressure, vibration, and speed. The acquisition time should cover different operating stages and working conditions of the pump to ensure the diversity and representativeness of the data.

[0123] Store the collected data in the database and create a historical data archive. Regularly organize and back up the data to ensure data security and integrity.

[0124] 2. Digital twin model training and optimization

[0125] The digital twin model of the turbomolecular pump is trained and optimized using the collected historical data. The model parameters and structure are adjusted to accurately simulate the actual operating behavior of the pump.

[0126] During the model training process, we use methods such as cross-validation and error analysis to evaluate the performance and accuracy of the model. We continuously improve the model training algorithm and strategy to improve the model's fit and generalization ability.

[0127] 3. Fault prediction model training and verification

[0128] Based on historical fault data and real-time operation data, the fault prediction model is trained. The fault data is divided into a training set and a test set. The model is trained with the training set, and then the prediction effect of the model is verified with the test set.

[0129] By adjusting the model's parameters and algorithms, the performance of the fault prediction model is optimized to ensure that the model can accurately identify and predict different types of faults.

[0130] (III) System operation and maintenance

[0131] 1. System startup and initialization

[0132] After the system is installed, start and initialize the system. Check the operating status of the hardware and software systems to ensure that the communication between the modules is normal.

[0133] The digital twin model and fault prediction model were loaded, the system was set to real-time monitoring mode, and the operating status of multiple turbomolecular pumps began to be monitored and analyzed.

[0134] 2. Real-time monitoring and data analysis

[0135] The system collects the operating data of multiple pumps in real time, and performs preprocessing and analysis. According to the health status evaluation index system, the health index value of each pump is calculated to monitor the health status of the pump in real time.

[0136] 3. Maintain decision execution and feedback

[0137] (1) Key component life prediction service: predict the future operating status of the pump and its key components. First, formulate a maintenance plan based on the prediction results, regularly inspect and maintain the pump, replace worn parts, and ensure the normal operation of the equipment. Second, if an abnormal situation is found or a possible failure is predicted, an early warning signal is issued in time, and relevant information and maintenance decisions are pushed to the user interface.

[0138] The user arranges maintenance personnel to maintain the corresponding turbomolecular pump according to the maintenance decision suggestions provided by the system. During the maintenance process, the maintenance content, maintenance time, and replaced spare parts and other information are recorded.

[0139] After the maintenance is completed, the pump is tested and data is monitored to evaluate the maintenance effect. The post-maintenance operation data is fed back to the system to update the digital twin model and fault prediction model, and continuously optimize the performance and accuracy of the system.

[0140] (2) Remote optimization control service: Develop intelligent control strategies based on the digital twin model and real-time operation data. The controller automatically adjusts the pump speed, power and other parameters to achieve optimal pumping efficiency and performance.

[0141] (IV) Example Analysis

[0142] Taking a semiconductor manufacturing plant as an example, the plant has multiple production lines, each of which is equipped with multiple turbomolecular pumps. After adopting the fault prediction method for turbomolecular pumps of the embodiment of the present invention, real-time monitoring and health management of all turbomolecular pumps are achieved.

[0143] Health status assessment and fault warning: Through the health status assessment function of the system, the abnormal temperature rise and increased vibration amplitude of one of the pumps were discovered in time. The system issued a fault warning signal based on the set health indicator threshold and indicated that the possible cause of the fault was bearing wear.

[0144] Maintenance decision making and implementation: Based on the maintenance decision suggestions provided by the system, the factory immediately arranged for maintenance personnel to shut down the pump for maintenance. After inspection, the maintenance personnel found that the bearing was indeed severely worn, so they replaced the bearing in time and inspected and maintained other parts of the pump.

[0145] Effect evaluation: After maintenance, the pump's operating status returned to normal, and the temperature and vibration indicators returned to normal range. By comparing and analyzing the data before and after maintenance, it was found that this method effectively reduced equipment downtime, improved production efficiency, and reduced maintenance costs.

[0146] In addition, since the system can monitor the operating status of multiple pumps in real time, it can detect potential faults in advance and avoid production line shutdowns caused by equipment failures, providing strong guarantees for the factory's production operations.

[0147] In this embodiment, a fault prediction device for a turbomolecular pump is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0148] This embodiment provides a fault prediction device for a turbomolecular pump, such as Figure 4 As shown, including:

[0149] The acquisition module 401 is used to acquire the operation data of the turbomolecular pump according to the set sampling frequency, and transmit the operation data through the data transmission network. The operation data includes historical operation data and real-time operation data.

[0150] The data processing module 402 is used to perform data cleaning, denoising, normalization and format conversion on the historical operation data.

[0151] The judgment module 403 is used to calculate the health index value of the turbomolecular pump according to the processed historical operation data, and judge whether the turbomolecular pump fails according to the health index value to obtain historical fault data.

[0152] The prediction module 404 is used to establish a fault prediction model based on historical fault data and real-time operation data, and use the fault prediction model to predict the fault state of the turbomolecular pump.

[0153] In some optional embodiments, the device further comprises:

[0154] The marking module is used to timestamp the running data.

[0155] In some optional implementations, the determination module 403 includes:

[0156] The setting unit is used to set the threshold range of each health indicator of the turbomolecular pump.

[0157] The first establishing unit is used to establish a digital twin model of the turbomolecular pump by using three-dimensional modeling technology and simulation technology according to the threshold range of each health indicator. The health indicators include temperature health indicators, pressure health indicators, vibration health indicators and performance efficiency indicators.

[0158] A computing unit is used to calculate the health index value of the turbomolecular pump using the digital twin model.

[0159] The determination unit is configured to determine that a fault occurs in the turbomolecular pump if the health indicator value exceeds a threshold range of the corresponding health indicator.

[0160] In some optional implementations, the prediction module 404 includes:

[0161] The second establishment unit is used to establish a fault prediction model of the turbomolecular pump by using machine learning algorithms and data analysis technology.

[0162] The partitioning unit is used to partition the historical fault data and the real-time operation data into a training data set and a test data set according to a preset ratio.

[0163] The training unit is used to train the fault prediction model using the training data set.

[0164] The verification unit is used to verify the fault prediction model using a test data set.

[0165] In some optional implementations, the verification unit includes:

[0166] The verification subunit is used to verify the generalization ability and effectiveness of the fault prediction model through the cross-validation method.

[0167] In some optional embodiments, the device further comprises:

[0168] The early warning module is used to send out an early warning signal when the fault state prediction result of the turbomolecular pump is a fault. The early warning signal includes the fault type, occurrence time and possibility.

[0169] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0170] The fault prediction device for the turbomolecular pump in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0171] The embodiment of the present invention also provides a computer device having the above Figure 4 A fault prediction device for a turbomolecular pump is shown.

[0172] See also Figure 5 , Figure 5 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 5 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 5 A processor 10 is taken as an example.

[0173] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0174] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0175] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0176] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0177] The computer device also includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30 and the output device 40 may be connected via a bus or other means. Figure 5 The example of connecting through bus is taken in the following.

[0178] The input device 30 can receive input digital or character information and generate key signal input related to user settings and function control of the computer device, such as a touch screen, etc. The output device 40 can include a display device, etc.

[0179] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0180] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0181] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope of the present application.

Claims

1. A fault prediction method for a turbomolecular pump, characterized in that: The method comprises: Collecting the operating data of the turbomolecular pump according to the set sampling frequency, and transmitting the operating data through the data transmission network, wherein the operating data includes historical operating data and real-time operating data; Performing data cleaning, denoising, normalization and format conversion on the historical operation data; Calculating a health index value of the turbomolecular pump according to the processed historical operation data, and determining whether the turbomolecular pump fails according to the health index value to obtain historical fault data; A fault prediction model is established based on historical fault data and real-time operation data, and the fault prediction model is used to predict the fault state of the turbomolecular pump.

2. The method according to claim 1, characterized in that: After collecting the operating data of the turbomolecular pump according to the set sampling frequency, the method further includes: The operating data is time-stamped.

3. The method according to claim 1, characterized in that The determining whether the turbomolecular pump fails according to the health indicator value includes: Set the threshold range of each health indicator of the turbomolecular pump; According to the threshold range of each health indicator, a digital twin model of the turbomolecular pump is established using three-dimensional modeling technology and simulation technology, wherein the health indicators include a temperature health indicator, a pressure health indicator, a vibration health indicator, and a performance efficiency indicator; Use digital twin models to calculate health indicators of turbomolecular pumps; If the health indicator value exceeds a threshold range of a corresponding health indicator, it is determined that the turbomolecular pump fails.

4. The method according to claim 1, characterized in that The method of establishing a fault prediction model based on historical fault data and real-time operation data includes: Use machine learning algorithms and data analysis techniques to establish a fault prediction model for turbomolecular pumps; Divide historical fault data and real-time operation data into training data sets and test data sets according to a preset ratio; Using the training data set to train the fault prediction model; The fault prediction model is verified using the test data set.

5. The method according to claim 4, characterized in that The verifying the fault prediction model by using the test data set includes: The generalization ability and effectiveness of the fault prediction model are verified by cross-validation method.

6. The method according to claim 1, characterized in that After predicting the fault state of the turbomolecular pump using the fault prediction model, the method further includes: When the fault state prediction result of the turbomolecular pump is that a fault occurs, an early warning signal is issued, and the early warning signal includes the fault type, occurrence time and possibility.

7. A fault prediction device for a turbomolecular pump, characterized in that: The device comprises: A collection module, used to collect the operation data of the turbomolecular pump according to a set sampling frequency, and transmit the operation data through a data transmission network, wherein the operation data includes historical operation data and real-time operation data; A data processing module, used for performing data cleaning, denoising, normalization and format conversion on the historical operation data; A judgment module, used to calculate a health index value of the turbomolecular pump according to the processed historical operation data, and judge whether a fault occurs in the turbomolecular pump according to the health index value to obtain historical fault data; The prediction module is used to establish a fault prediction model based on historical fault data and real-time operation data, and use the fault prediction model to predict the fault state of the turbomolecular pump.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the fault prediction method for a turbomolecular pump according to any one of claims 1 to 6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the fault prediction method for a turbomolecular pump according to any one of claims 1 to 6.

10. A computer program product, characterized in that The method comprises computer instructions for causing a computer to execute the fault prediction method for a turbomolecular pump according to any one of claims 1 to 6.

Images