Intelligent twinborn model construction method and system for process industry

Abstract

The invention discloses a process industry intelligent twin model construction method and system. The objective of the invention is to solve the problem of crash or paralysis of a training model caused by huge data in the prior art. Comprising the following steps: firstly, acquiring and storing data of equipment in the process industry, performing classification management on the data, secondly, constructing an equipment data model through a neural network model, and performing training based on the data subjected to classification management to form an intelligent twin model. And finally, performing efficient processing and analysis on newly input equipment data by utilizing the trained intelligent twin model, monitoring the operation state of the equipment in real time, and ensuring stable operation of the system. According to the method, mass data generated by equipment in the process industry can be effectively processed.

Description

Technical Field

[0001] The present invention relates to the technical field of twin model establishment, and particularly relates to a method and system for constructing an intelligent twin model in the process industry. Background Art

[0002] With the continuous development of technology, the requirements for refined industrial production monitoring are getting higher and higher. Traditional industrial production monitoring systems mainly rely on manual labor and mechanical equipment for operation. Workers need to conduct regular inspections, check the operating status of equipment, record relevant data, and handle any abnormal situations. In addition, factories also use various sensors and cameras to collect data and images during the production process to monitor the production process and equipment status.

[0003] There are some problems with this traditional monitoring method. First, the data in the equipment database is relatively complex. The main data includes video photos, dimensions, equipment models, equipment sensor data, and the operating data of equipment during production. When the huge amount of data is directly transmitted to the training model, especially in a large-scale and high-traffic production environment where the data is even more massive, it often causes the training model to freeze or crash. Therefore, we need to classify the data before data transmission.

[0004] In addition, no model is formed for continuous reinforcement to form a line system intelligent twin, making it difficult to achieve the expected goals of description, diagnosis, prediction, and early warning. Summary of the Invention

[0005] Based on this, the embodiments of the present application provide a method and system for constructing an intelligent twin model in the process industry to solve the problem that the huge amount of data in the prior art may cause the training model to freeze or crash.

[0006] In a first aspect, a method for constructing an intelligent twin model in the process industry is provided. The method includes:

[0007] Obtain and store various data of process industry equipment, and classify and manage the data; wherein, the data is classified into digital delivery data and operating data, and the digital delivery data includes structured data, unstructured data, and simple models, and the operating data includes sensor data and production data;

[0008] Construct an equipment data model through a neural network model, and train the equipment data model based on the classified and managed data to obtain an intelligent twin model;

[0009] Use the trained intelligent twin model to process and analyze newly input process industry equipment data.

[0010] Optionally, the digitally delivered data specifically includes the design parameters, specifications, and historical maintenance records of the device; the operating data specifically includes real-time sensor data and operating parameters during the production process.

[0011] Optionally, an equipment data model is constructed through a neural network model, including:

[0012] According to the application scenarios of process industry equipment data, select the corresponding neural network architecture; among them, the selected neural network structure includes at least multi-layer perceptron, convolutional neural network, recurrent neural network, and hybrid neural network;

[0013] Initialize the parameters of the selected neural network model and configure the hyperparameters of the model;

[0014] Construct an initial neural network model according to the selected neural network architecture and the configured hyperparameters; among them, it includes defining the structures of the input layer, hidden layer, and output layer, as well as the connection methods between layers.

[0015] Optionally, the equipment data model specifically includes an equipment data small model and an equipment data general large model. Based on the classified and managed data, the equipment data model is trained to obtain an intelligent twin model, specifically including:

[0016] Train the initial neural network model based on the classified and managed data to obtain an equipment data small model;

[0017] Train the equipment data small model based on the classified and managed data to obtain an equipment data general large model;

[0018] Perform data verification on the equipment data general large model, and use the verified equipment data general large model as the intelligent twin model.

[0019] Optionally, before classifying and managing the data, it also includes preprocessing the data, including removing duplicate descriptions of the received data to avoid data transmission errors.

[0020] Optionally, using the trained intelligent twin model to process and analyze newly input process industry equipment data also includes:

[0021] Determine the operating state of the process industry equipment under the set working conditions based on the newly input process industry equipment data; by analyzing the operating state data of the equipment, identify potential failure modes, and generate a warning signal when the model predicts that the operating state of the equipment may exceed the normal range or approach the failure threshold.

[0022] In a second aspect, a process industry intelligent twin model construction system is provided, and the system includes:

[0023] A classification management module is used to obtain and store various data of process industry equipment and classify and manage the data. Among them, the data is classified into digital delivery data and operation data. The digital delivery data includes structured data, unstructured data and simple models, and the operation data includes sensor data and production data;

[0024] A model training module is used to construct an equipment data model through a neural network model and train the equipment data model based on the classified and managed data to obtain an intelligent twin model;

[0025] A processing and analysis module is used to process and analyze newly input process industry equipment data by using the trained intelligent twin model.

[0026] In a third aspect, an electronic device is provided, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the method for constructing an intelligent twin model for the process industry described in any one of the first aspects above is implemented.

[0027] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method for constructing an intelligent twin model for the process industry described in any one of the first aspects above is implemented.

[0028] In a fifth aspect, a computer program product is provided, which includes a computer program / instructions. When the computer program / instructions are executed by a processor, the method for constructing an intelligent twin model for the process industry described in any one of the first aspects above is implemented.

[0029] The beneficial effects brought by the technical solutions provided in the embodiments of the present application at least include that through data classification management and the incremental learning, lightweight processing and batch processing mechanisms of the intelligent twin model, the present invention can effectively process the massive data generated by process industry equipment. Compared with the prior art, this method avoids the system crash or paralysis problem caused by directly inputting huge data into the training model, and significantly improves the stability and reliability of the system. Description of the Drawings

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, without creative efforts, other implementation drawings can also be obtained according to the provided drawings.

[0031] Figure 1 It is a step flow chart of a method for constructing an intelligent twin model for the process industry provided by the embodiment of the present application;

[0032] Figure 2 Block diagram of an intelligent twin model construction system for the process industry provided by an embodiment of the present application;

[0033] Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0034] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0035] In the description of the present invention, the terms "including", "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily limit to the clearly listed steps or units, but may also include other steps or units inherent to these processes, methods, products or devices although not clearly listed, or steps or units added based on further optimized solutions conceived from the present invention.

[0036] Please refer to Figure 1 , which shows a flowchart of an intelligent twin model construction method for the process industry provided by an embodiment of the present application, and may include the following steps:

[0037] S1. Obtain and store various data of process industry equipment, and classify and manage the data.

[0038] Among them, the specific classification includes digital delivery data and operation data. The digital delivery data includes structured data, unstructured data and simple models. The operation data includes sensor data and production data.

[0039] In this step, the operation data of process industry equipment is collected through a sensor network, an industrial control system (such as DCS, PLC), an equipment management system, etc. The digital delivery data is obtained from an enterprise resource planning (ERP) system, equipment design documents, historical maintenance records.

[0040] Among them, the types of collected data include:

[0041] Structured data: such as equipment parameters, operation indicators, production data, etc.

[0042] Unstructured data: such as equipment images, videos, log files, etc.

[0043] Simple models: such as geometric models of equipment, flowcharts, etc.

[0044] Store the collected data in a distributed storage system (such as Hadoop, Ceph) or cloud storage to support the storage and access of large-scale data. For data with high real-time requirements (such as sensor data), a time series database (such as InfluxDB) can be used for storage.

[0045] Data classification management specifically includes removing noise data, filling in missing values, and correcting incorrect data. Convert different types of data to the same scale for subsequent processing. Classify the data into categories such as structured data, unstructured data, sensor data, and production data according to the type, source, and use of the data. Establish an index for the stored data to facilitate fast retrieval and access. Ensure data security, perform regular backups to prevent data loss or damage.

[0046] S2. Construct a device data model through a neural network model, and perform model training on the device data model based on the classified and managed data to obtain an intelligent twin model.

[0047] In this step, select a suitable neural network architecture according to the characteristics and application scenarios of the device data, such as multi-layer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN) and its variants (LSTM, GRU).

[0048] For time series data (such as sensor data), preferably choose RNN or LSTM; for image data, choose CNN.

[0049] Use a random initialization method (such as Xavier initialization, He initialization) to initialize the weight and bias parameters. If there is a pre-trained model in the relevant field, its parameters can be used as the initial parameters to accelerate the training process and improve the model performance. Determine hyperparameters such as the learning rate, hidden layer size, activation function, regularization parameter, and optimizer. Select an appropriate batch size and number of training epochs according to the data volume and model complexity. Define the structure of the input layer, hidden layer, and output layer, as well as the connection method between each layer. For multi-modal data (such as containing both image and sensor data at the same time), a hybrid neural network architecture can be constructed.

[0050] Use the classified and managed data as the training set, calculate the output result through forward propagation, and use a loss function (such as mean square error, cross-entropy loss) to evaluate the model performance. Update the model parameters through backpropagation, and repeat the training process until the model converges or reaches the preset number of training epochs.

[0051] Evaluate the performance of the model with the validation set to ensure the accuracy and generalization ability of the model. Adjust the hyperparameters or model structure according to the validation results to optimize the model performance.

[0052] S3. Use the trained intelligent twin model to process and analyze newly input process industry equipment data.

[0053] Perform preprocessing operations such as data cleaning and normalization on the newly input data to ensure that the data format is consistent with that during training. Input the preprocessed data into the trained intelligent twin model for real-time inference.

[0054] The model outputs the operating status of the equipment, including performance indicators, real-time values of key parameters, etc. Compare the operating status output by the model with the set normal operating range to evaluate whether the equipment is in a normal operating state.

[0055] If an anomaly is detected, further analyze the type and cause of the anomaly. The intelligent twin model dynamically updates the model parameters through incremental learning to adapt to the changes in new data. The model monitors the changes in data characteristics in real time, automatically adjusts the model structure or parameters to ensure the accuracy and stability of the model. Batch process the newly input large amount of data to avoid exhausting system resources due to loading too much data at one time. Use lightweight technologies such as model pruning and quantization to reduce the computational complexity and storage requirements of the model.

[0056] During the operation of the model, monitor the usage of system resources in real time and dynamically adjust the allocation of computing resources. When the data volume is large, automatically allocate more computing resources to ensure the normal operation of the model. The model automatically evaluates the performance indicators (such as prediction accuracy, response time) after processing the data and makes optimization adjustments according to the evaluation results. Record the process of processing the data and the performance feedback to provide a reference for subsequent model optimization.

[0057] Please refer to Figure 2 , which shows a block diagram of a process industry intelligent twin model construction system provided by an embodiment of the present application. As Figure 2 shown, the system may include:

[0058] A classification management module for acquiring and storing various data of process industry equipment and classifying and managing the data; among them, the data is classified into digital delivery data and operation data, and the operation data includes sensor data and production data; the digital delivery data includes structured data, unstructured data and simple models.

[0059] A model training module for constructing an equipment data model through a neural network model and training the equipment data model based on the data after classification management to obtain an intelligent twin model.

[0060] A processing and analysis module for using the trained intelligent twin model to process and analyze newly input process industry equipment data.

[0061] For the specific limitations of the intelligent twin model construction system in the process industry, reference can be made to the limitations of the intelligent twin model construction method in the process industry described above, which will not be elaborated here. Each module in the above-mentioned intelligent twin model construction system in the process industry can be implemented in whole or in part by software, hardware, and their combinations. Each of the above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory of the computer device in the form of software, so as to facilitate the processor to call and execute the operations corresponding to each of the above modules.

[0062] In one embodiment, an electronic device is provided. The electronic device can be a computer, and its internal structure diagram can be as Figure 3 shown. The electronic device includes a processor, a memory, and a network interface connected through a system bus. Among them, the processor of the device is used to provide computing and control capabilities. The memory of the 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 database of the computer device is used for the data of the intelligent twin model construction in the process industry. 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, it implements an intelligent twin model construction method in the process industry.

[0063] Those skilled in the art can understand that the structure shown in Figure 3 is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0064] In one embodiment, a computer-readable storage medium is further provided, on which a computer program is stored, covering all or part of the processes in the method of the above embodiment.

[0065] In one embodiment, a computer program product is further provided, including a computer program / instructions, covering all or part of the processes in the method of the above embodiment.

[0066] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. 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 methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in M forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Symchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0067] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0068] The above embodiments only represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. A method for constructing an intelligent twin model for a process industry, characterized in that: The method comprises: Acquire and store various data of process industry equipment, and classify and manage the data; the data is classified into digital delivery data and operation data, the digital delivery data includes structured data, unstructured data and simple models, and the operation data includes sensor data and production data; A device data model is constructed through a neural network model, and a model training is performed on the device data model based on the classified and managed data to obtain an intelligent twin model; Use the trained intelligent twin model to process and analyze newly input process industry equipment data.

2. The method for constructing an intelligent twin model for a process industry according to claim 1, characterized in that: Digital delivery data specifically includes equipment design parameters, specifications, and historical maintenance records; operation data specifically includes real-time sensor data and operating parameters during the production process.

3. The method for constructing an intelligent twin model for a process industry according to claim 1, characterized in that: Build a device data model through a neural network model, including: According to the application scenario of process industry equipment data, select the corresponding neural network architecture; the selected neural network structure includes at least multi-layer perceptron, convolutional neural network, recurrent neural network and hybrid neural network; Initialize the parameters of the selected neural network model and configure the model's hyperparameters; Based on the selected neural network architecture and configured hyperparameters, an initial neural network model is constructed, including defining the structure of the input layer, hidden layer, and output layer, as well as the connection method between the layers.

4. The method for constructing an intelligent twin model for a process industry according to claim 3, characterized in that: The device data model specifically includes a small device data model and a general large device data model. The device data model is trained based on the classified and managed data to obtain an intelligent twin model, which specifically includes: The initial neural network model is trained based on the classified and managed data to obtain a small model of equipment data; Based on the classified and managed data, the small model of equipment data is trained to obtain a general large model of equipment data; The universal big model of equipment data is verified, and the verified universal big model of equipment data is used as the intelligent twin model.

5. The method for constructing an intelligent twin model for a process industry according to claim 1, characterized in that: Before classifying and managing the data, it also includes preprocessing the data, including repeating the description of the received data to avoid data transmission errors.

6. The method for constructing an intelligent twin model for a process industry according to claim 1, characterized in that: Using the trained intelligent twin model to process and analyze newly input process industry equipment data also includes: Determine the operating status of process industry equipment under set working conditions based on newly input process industry equipment data; identify potential failure modes by analyzing the equipment's operating status data; and generate a warning signal when the model predicts that the equipment's operating status may exceed the normal range or approach the failure threshold.

7. A process industry intelligent twin model construction system, characterized in that: The system comprises: The classification management module is used to obtain and store various data of process industry equipment and classify and manage the data; the data is classified into digital delivery data and operation data, the digital delivery data includes structured data, unstructured data and simple models, and the operation data includes sensor data and production data; A model training module is used to construct a device data model through a neural network model, and to perform model training on the device data model based on the classified and managed data to obtain an intelligent twin model; The processing and analysis module is used to process and analyze newly input process industry equipment data using the trained intelligent twin model.

8. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images