Machine learning model automatic generation and deployment method based on process simulation platform

Abstract

The invention discloses a machine learning model automatic generation and deployment method based on a process simulation platform, and the method comprises the steps: firstly collecting model training data, and carrying out the preprocessing of the collected model training data; then, a to-be-trained machine learning model is newly built based on a pre-built automatic machine learning model framework, automatic model training is carried out on the newly built machine learning model based on the preprocessed model training data, and a data model completing model training is generated; and finally, automatically deploying the generated data model to a process simulation platform so as to connect a mechanism model on the process simulation platform to carry out process calculation and prediction.

Description

Technical Field

[0001] The present invention relates to the field of industrial process modeling, and in particular to a method for automatically generating and deploying a machine learning model based on a process simulation platform. Background Art

[0002] In modern industrial production, process simulation software is widely used to model and optimize complex industrial processes. However, traditional mechanism-based modeling methods rely on expert experience and a large amount of experimental data. The modeling process is cumbersome and time-consuming, especially when the process is complex and there are many variables. It is difficult to build a high-precision model in a short time. In addition, as process conditions change, mechanism models also need to be frequently readjusted and verified, which increases the complexity and cost of model maintenance. Therefore, industrial process modeling based on data-driven and machine learning technology has gradually become a trend.

[0003] Although existing automatic machine learning technologies can alleviate the modeling burden to a certain extent, they still have some limitations. For example, most automatic machine learning tools cannot be directly integrated with process simulation platforms, and they do not provide sufficient support for complex multi-input and multi-output industrial processes, making it difficult to promote and apply them on a large scale in actual production environments. Therefore, there is an urgent need for a complete solution with high integration that can automatically collect data, generate models, and deploy them to achieve efficient modeling and optimization of industrial processes. Summary of the invention

[0004] A brief summary of one or more aspects is given below to provide a basic understanding of these aspects. This summary is not an exhaustive overview of all conceived aspects, and is neither intended to identify the key or critical elements of all aspects nor to define the scope of any or all aspects. Its only purpose is to give some concepts of one or more aspects in a simplified form as a prelude to a more detailed description that will be given later.

[0005] The purpose of the present invention is to solve the above problems and provide a method for automatic generation and deployment of machine learning models based on a process simulation platform. A pre-built automatic machine learning model framework is used to create a new machine learning model to be trained. Through automated data collection, modeling, training and deployment, manual intervention is greatly reduced, the modeling cycle is shortened, and the modeling efficiency and accuracy are improved.

[0006] The technical solution of the present invention is:

[0007] The present invention provides a method for automatically generating and deploying a machine learning model based on a process simulation platform, comprising the following steps:

[0008] Step S1: Collect model training data and preprocess the collected model training data;

[0009] Step S2: creating a new machine learning model to be trained based on the pre-built automatic machine learning model framework;

[0010] Step S3: Automatically train the newly created machine learning model based on the preprocessed model training data to generate a data model that completes the model training;

[0011] Step S4: Automatically deploy the generated data model to the process simulation platform, thereby connecting the mechanism model on the process simulation platform to calculate and predict the process.

[0012] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, in step S1, the method for automatic generation and deployment of machine learning models based on a process simulation platform collects key process parameters during the process operation process based on the process simulation platform or the factory real-time database as model training data, and then preprocesses the collected model training data; wherein the preprocessing includes data cleaning, normalization, missing value completion and feature selection.

[0013] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, in step S1, the method for automatic generation and deployment of machine learning models based on a process simulation platform preprocesses the collected model training data, and then specifies the input variable feature columns and the output variable, i.e., the label columns, for the collected model training data, thereby obtaining tabular model training data.

[0014] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, in step S2, the method for automatic generation and deployment of machine learning models based on a process simulation platform uses AutoGluon to pre-build an automatic machine learning model framework and deploy it in the process simulation software, and then constructs and trains the machine learning model based on the constructed automatic machine learning model framework; wherein the automatic machine learning model framework supports the configuration of multiple model inputs and multiple model outputs.

[0015] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, after a new machine learning model is created, the method for automatic generation and deployment of machine learning models based on a process simulation platform automatically generates a front-end interactive interface, and automatically trains the machine learning model through the generated front-end interactive interface; wherein the front-end interactive interface includes a parameter configuration area, a model selection control area, a training process monitoring area, and a prediction result visualization area.

[0016] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, when the method for automatic generation and deployment of machine learning models based on a process simulation platform performs automatic model training on the machine learning model through a front-end interactive interface, the user configures and dynamically adjusts the model input and model output of the machine learning model through the parameter configuration area, selects the required machine learning model type in the model selection control area, and then monitors the model training process through the training process monitoring area, and finally views the model training prediction results through the prediction result visualization area.

[0017] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, automatic model training includes model selection, feature engineering and hyperparameter tuning; wherein, when the machine learning model is automatically generated and deployed to perform automatic model training on the machine learning model, the model training process is accelerated by parallel computing, and the Bayesian optimization method is used for hyperparameter tuning.

[0018] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, the method for automatic generation and deployment of machine learning models based on a process simulation platform performs data transmission and model deployment through an API interface; wherein, when the method for automatic generation and deployment of machine learning models based on a process simulation platform performs automatic model training, the model training data is updated in real time for the machine learning model through the API interface, and the generated data model is automatically pushed to the process simulation platform for deployment through the API interface, thereby connecting the mechanism model on the process simulation platform to calculate and predict the process.

[0019] According to an embodiment of the method for automatically generating and deploying a machine learning model based on a process simulation platform of the present invention, after generating a data model, the method for automatically generating and deploying a machine learning model based on a process simulation platform automatically monitors the predictability of the generated data model based on preset performance indicators, thereby determining whether to re-train the model; wherein,

[0020] If the monitored predictability is lower than the preset performance index, model retraining and parameter adjustment are automatically triggered;

[0021] If the monitored predictability is greater than or equal to the preset performance indicator, the generated data model will be automatically deployed to the process simulation platform.

[0022] According to an embodiment of the method for automatic generation and deployment of machine learning models based on a process simulation platform of the present invention, the method for automatic generation and deployment of machine learning models based on a process simulation platform also sets an operation status monitoring strategy, which monitors whether the operation status of the deployed data model is abnormal; if so, perform model maintenance; if not, continue monitoring.

[0023] The present invention also provides a computer readable medium storing a computer program code, wherein the computer program code implements the method as described above when executed by a processor.

[0024] The present invention also provides a device for automatically generating and deploying a machine learning model based on a process simulation platform, comprising:

[0025] a memory for storing instructions executable by a processor; and

[0026] A processor is used to execute the instructions to implement the method as described above.

[0027] Compared with the prior art, the present invention has the following beneficial effects: the present invention aims at the automated generation and deployment of machine learning for the process simulation platform, and uses a pre-built automatic machine learning model framework to create a new machine learning model to be trained. Through automated data collection, modeling, training and deployment, it not only greatly shortens the modeling cycle, but also reduces manual intervention, thereby improving modeling efficiency and accuracy. In addition, the present invention supports the configuration of multiple model inputs and model outputs, so it is suitable for complex multi-input and multi-output industrial process modeling, meeting the multi-objective control requirements of the chemical, petrochemical, pharmaceutical and other industries. At the same time, the present invention can also seamlessly integrate the generated data model with the process simulation platform, thereby realizing an integrated solution from data collection to real-time prediction, effectively improving the control accuracy of the process, reducing energy consumption, and improving product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The above features and advantages of the present invention can be better understood after reading the detailed description of the embodiments of the present disclosure in conjunction with the following drawings. In the drawings, the components are not necessarily drawn to scale, and components with similar related properties or features may have the same or similar reference numerals.

[0029] Figure 1 It is a flowchart showing an embodiment of a method for automatic generation and deployment of a machine learning model based on a process simulation platform of the present invention. DETAILED DESCRIPTION

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. Unless it is obvious from the language environment or otherwise explained, the same reference numerals in the figures represent the same structure or operation.

[0031] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0032] Unless otherwise specifically stated, the relative arrangement, numerical expressions and numerical values ​​of the parts and steps set forth in these embodiments do not limit the scope of the present application. At the same time, it should be understood that, for ease of description, the sizes of the various parts shown in the accompanying drawings are not drawn according to the actual proportional relationship. The technology, method and equipment known to those of ordinary skill in the relevant field may not be discussed in detail, but in appropriate cases, the technology, method and equipment should be considered as a part of the authorization specification. In all examples shown and discussed here, any specific value should be interpreted as being merely exemplary, rather than as a limitation. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters represent similar items in the following drawings, so that once a certain item is defined in an accompanying drawing, it does not need to be further discussed in subsequent drawings.

[0033] Disclosed herein is an embodiment of a method for automatically generating and deploying a machine learning model based on a process simulation platform. Figure 1 is a flow chart showing an embodiment of a method for automatically generating and deploying a machine learning model based on a process simulation platform of the present invention. Figure 1 ,The following is a detailed description of the steps of the automatic generation and deployment method of machine learning models based on the process simulation platform.

[0034] Step S1: Collect model training data and preprocess the collected model training data.

[0035] In this embodiment, before training the machine learning model, key process parameters during the process operation are first collected from the process simulation platform or the factory real-time database as model training data, and then the collected model training data is preprocessed; wherein the preprocessing includes data cleaning, normalization, missing value incompleteness and feature selection.

[0036] Specifically, in this embodiment, in order to obtain model training data in a timely manner, a scheduled task or data collection trigger condition can be set to collect key process parameters such as temperature, pressure, flow rate, component concentration, etc. during the process operation from the process simulation platform or the factory real-time database as model training data to ensure the real-time and continuity of the data. At the same time, in order to ensure the validity of the data, the noise and outliers are removed by cleaning the collected data, and then the variables of different dimensions are converted to the same scale through normalization, and the missing values ​​are supplemented by interpolation, mean filling and other methods to ensure the integrity of the data. Finally, feature selection is also performed on the collected data to screen out features that have a significant impact on the model.

[0037] In addition, in this embodiment, after the preprocessing of the model training data is completed through the above steps, a feature column and a label column are formulated for the preprocessed model training data to obtain a tabular model training data. Among them, the feature column represents the model input, and the label column represents the model output. The input features of the model are specified by constructing the tabular model training data.

[0038] Step S2: Create a new machine learning model to be trained based on the pre-built automatic machine learning model framework.

[0039] In this embodiment, AutoGluon is used to pre-build an automatic machine learning model framework, and then the machine learning model is built and trained based on the built automatic machine learning model framework. Among them, the automatic machine learning model framework supports the configuration of multiple model inputs and multiple model outputs.

[0040] Specifically, AutoGluon is an open source tool that can automatically perform model selection, feature engineering, and hyperparameter optimization, and supports complex industrial process modeling with multiple inputs and outputs. In the process of building a machine learning model, feature selection is performed through automated means to select features that have a greater impact on the target variable to improve the prediction accuracy of the model. Among them, the types of machine learning models supported by AutoGluon include regression models, classification models, and time series prediction models, etc. The specific selection depends on the characteristics of the industrial process.

[0041] In addition, in this embodiment, the model configuration supports multiple types of model inputs and model outputs, including numerical data, categorical data, and time series data. By supporting multiple inputs and multiple outputs, the present invention can effectively respond to the multi-variable and multi-objective modeling requirements in industrial processes. For example, in chemical production, when multiple input variables (such as temperature and pressure) affect multiple output targets (such as output and quality), the machine learning model established by AutoGluon can automatically adapt to this complex input-output relationship.

[0042] In this embodiment, after a new machine learning model to be trained is created based on the automatic machine learning model framework, a front-end interactive interface is automatically generated, and the machine learning model is automatically trained through the generated front-end interactive interface. Among them, the front-end interactive interface includes a parameter configuration area, a model selection control area, a training process monitoring area, and a prediction result visualization area. Therefore, when the machine learning model is automatically trained through the front-end interactive interface, the user can configure and dynamically adjust the model input and model output of the machine learning model through the parameter configuration area, and select the required machine learning model type in the model selection control area, and then monitor the model training process through the training process monitoring area, and finally view the model training prediction results through the prediction result visualization area.

[0043] In addition, in this embodiment, the front-end interactive interface also supports graphical display, such as time series graphs, pie charts, bar charts, etc., so that the prediction results can be displayed more intuitively and easily understood. Users can fine-tune the input parameters of the model through the interface and view the prediction results of the model in real time, thereby better supporting decision-making in the production process.

[0044] Step S3: Automatically train the newly created machine learning model based on the preprocessed model training data to generate a data model that completes the model training.

[0045] In this embodiment, automatic model training includes model selection, feature engineering and hyperparameter tuning. When the machine learning model is automatically generated and deployed, the model training process is accelerated by parallel computing, and the Bayesian optimization method is used to perform hyperparameter tuning.

[0046] Specifically, in this embodiment, when training the machine learning model, a suitable initial model framework is selected by model selection such as stacking and weighted average strategy, and then the most representative features are extracted from the original data by feature engineering such as feature extraction and feature selection to reduce the interference of redundant data on the model. At the same time, when training the model, the training process is accelerated by parallel computing, so that the training and verification of the model can be completed quickly, saving time. And for the parameter tuning of the model, Bayesian optimization is used to search the parameter space of the model, and the best parameter combination is automatically found, so as to optimize the hyperparameters of the machine learning model. In addition, in this embodiment, for large-scale data sets, distributed computing is also supported to ensure the efficiency of model training. At the same time, in this embodiment, multiple models can be set according to different needs to be trained separately, and then the prediction results of multiple models are integrated through model integration to improve the overall prediction accuracy and robustness.

[0047] Step S4: Automatically deploy the generated data model to the process simulation platform, thereby connecting the mechanism model on the process simulation platform to calculate and predict the process.

[0048] In this embodiment, after the required data model is generated through the above steps, data transmission and model deployment are performed through the API interface. Among them, when performing automatic model training, the model training data can also be updated in real time for the machine learning model through the API interface, and the generated data model can be automatically pushed to the process simulation platform for deployment through the API interface, and seamlessly integrated with the existing mechanism model, so as to connect the mechanism model on the process simulation platform to receive process parameters in real time and perform online prediction. By combining the real-time and flexibility of the data model, and the process physical constraints and interpretability of the mechanism model, the prediction accuracy is effectively improved and the reliability of the system is enhanced.

[0049] At the same time, in this embodiment, in addition to the above-mentioned real-time mode, the deployment and prediction of the data model can also adopt an offline mode. In the offline mode, prediction and analysis are performed based on historical data, thereby providing a basis for process optimization and improvement. In addition, this embodiment can also use historically stored molecular sieves to synthesize experimental data, establish a prediction model from operating conditions to molecular sieve types, and predict the types of molecular sieves produced by the experiment through the established prediction model, so as to provide guidance for the actual molecular sieve synthesis experiment, thereby making targeted adjustments to the experimental operating conditions and reducing the cost of R&D experiments.

[0050] In addition, in this embodiment, when calculating and predicting the model, anomaly detection and processing can also be performed on the prediction process of the model. When an abnormal situation is detected during the prediction process (such as a significant deviation between the model prediction result and the actual observation value), corresponding measures can be taken, such as notifying relevant maintenance personnel or rolling back to the previous stable model version, to ensure the safety and reliability of the production process.

[0051] Specifically, in this embodiment, after the data model is generated through the above steps, the predictability of the generated data model is automatically monitored according to the preset performance indicators to determine whether to re-train the model. Among them, if the monitored predictability is lower than the preset performance indicator, the model retraining and parameter adjustment are automatically triggered. If the monitored predictability is greater than or equal to the preset performance indicator, the generated data model is automatically deployed directly to the process simulation platform. In addition, this embodiment also sets an operation status monitoring strategy, which monitors whether the operation status of the deployed data model is abnormal. If so, perform model maintenance. If not, continue monitoring.

[0052] In this embodiment, in order to facilitate the maintenance and upgrading of the model, the model version management is also supported. After each model update and retraining, a new model version will be generated and stored, and the user can switch between different versions to ensure the stability of the production process.

[0053] The present invention utilizes historically stored molecular sieve synthesis experimental data to establish a prediction model from operating conditions to molecular sieve types, predicts the type of molecular sieve produced by the experiment, provides guidance for actual molecular sieve synthesis experiments, adjusts experimental operating conditions in a targeted manner, and reduces R&D experimental costs.

[0054] This specification also provides a computer-readable medium storing computer program code, which, when executed by a processor, implements the method for automatically generating and deploying a machine learning model based on a process simulation platform as described above.

[0055] This specification also provides a method for automatically generating and deploying a machine learning model based on a process simulation platform, including a memory for storing instructions executable by a processor, and a processor for executing instructions in the instruction memory to implement the method for automatically generating and deploying a machine learning model based on a process simulation platform as described above.

[0056] The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples and designs described herein, but should be granted the widest scope consistent with the principles and novel features disclosed herein.

[0057] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or a combination of the two. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps are generally described above in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. The technician may implement the described functionality in different ways for each specific application, but such implementation decisions should not be interpreted as resulting in a departure from the scope of the present invention.

[0058] The various illustrative logic blocks, modules, and circuits described in conjunction with the embodiments disclosed herein may be implemented or performed with a general purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in cooperation with a DSP core, or any other such configuration.

[0059] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software module may reside in a RAM memory, a flash memory, a ROM memory, an EPROM memory, an EEPROM memory, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor so that the processor can read and write information from / to the storage medium. In an alternative, a storage medium may be integrated into a processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and the storage medium may reside in a user terminal as discrete components.

[0060] In one or more exemplary embodiments, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented as a computer program product in software, each function may be stored on or transmitted by a computer-readable medium as one or more instructions or codes. Computer-readable media include both computer storage media and communication media, including any medium that facilitates the transfer of a computer program from one place to another. Storage media may be any available medium that can be accessed by a computer. As an example and not limitation, such a computer-readable medium may include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of an instruction or data structure and can be accessed by a computer. Any connection is also properly referred to as a computer-readable medium. For example, if the software is transmitted from a website, a server, or other remote source using a coaxial cable, a fiber optic cable, a twisted pair, a digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwaves, the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwaves are included in the definition of the medium. Disk and disc as used herein include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, wherein disk often reproduces data magnetically, while disc reproduces data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

Claims

1. A method for automatically generating and deploying a machine learning model based on a process simulation platform, characterized in that: The following steps are involved: Step S1: Collect model training data and preprocess the collected model training data; Step S2: creating a new machine learning model to be trained based on the pre-built automatic machine learning model framework; Step S3: Automatically train the newly created machine learning model based on the preprocessed model training data to generate a data model that completes the model training; Step S4: Automatically deploy the generated data model to the process simulation platform, thereby connecting the mechanism model on the process simulation platform to calculate and predict the process.

2. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 1, characterized in that: In step S1, the automatic generation and deployment method of the machine learning model based on the process simulation platform collects the key process parameters during the process operation from the process simulation platform or the factory real-time database as model training data, and then preprocesses the collected model training data; wherein the preprocessing includes data cleaning, normalization, missing value completion and feature selection.

3. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 1, characterized in that: In step S1, the automatic generation and deployment method of machine learning models based on the process simulation platform preprocesses the collected model training data, and then specifies the input variable feature columns and the output variable, i.e., the label columns, for the collected model training data, thereby obtaining tabular model training data.

4. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 1, characterized in that: In step S2, the method for automatic generation and deployment of machine learning models based on a process simulation platform uses AutoGluon to pre-build an automatic machine learning model framework and deploy it in the process simulation software, and then builds and trains the machine learning model based on the constructed automatic machine learning model framework; wherein the automatic machine learning model framework supports the configuration of multiple model inputs and multiple model outputs.

5. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 4, characterized in that: After a new machine learning model is created, the method for automatically generating and deploying a machine learning model based on a process simulation platform automatically generates a front-end interactive interface, and automatically trains the machine learning model through the generated front-end interactive interface; wherein the front-end interactive interface includes a parameter configuration area, a model selection control area, a training process monitoring area, and a prediction result visualization area.

6. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 5, characterized in that: When the automatic generation and deployment method of machine learning models based on the process simulation platform performs automatic model training on the machine learning model through the front-end interactive interface, the user configures and dynamically adjusts the model input and model output of the machine learning model through the parameter configuration area, selects the required machine learning model type in the model selection control area, and then monitors the model training process through the training process monitoring area, and finally views the model training prediction results through the prediction result visualization area.

7. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 1, characterized in that: Automatic model training includes model selection, feature engineering and hyperparameter tuning; wherein, when the machine learning model is automatically generated and deployed for automatic model training of the machine learning model, the model training process is accelerated by parallel computing, and the Bayesian optimization method is used for hyperparameter tuning.

8. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 1, characterized in that: The method for automatically generating and deploying machine learning models based on a process simulation platform performs data transmission and model deployment through an API interface; wherein, when the method for automatically generating and deploying machine learning models based on a process simulation platform performs automatic model training, the model training data is updated in real time for the machine learning model through the API interface, and the generated data model is automatically pushed to the process simulation platform for deployment through the API interface, thereby connecting the mechanism model on the process simulation platform to calculate and predict the process.

9. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 4, characterized in that: After the data model is generated by the automatic generation and deployment method of the machine learning model based on the process simulation platform, the predictability of the generated data model is automatically monitored based on preset performance indicators, so as to determine whether to re-train the model; wherein, If the monitored predictability is lower than the preset performance index, model retraining and parameter adjustment are automatically triggered; If the monitored predictability is greater than or equal to the preset performance indicator, the generated data model will be automatically deployed to the process simulation platform.

10. The method for automatically generating and deploying a machine learning model based on a process simulation platform according to claim 4, characterized in that: The method for automatically generating and deploying machine learning models based on a process simulation platform also sets an operation status monitoring strategy, which monitors whether the operation status of the deployed data model is abnormal; if so, perform model maintenance; if not, continue monitoring.

11. A computer readable medium storing computer program code, characterized in that: The computer program code implements the method according to any one of claims 1 to 10 when executed by a processor.

12. A machine learning model automatic generation and deployment device, characterized in that: include: a memory for storing instructions executable by a processor; as well as A processor, configured to execute the instructions to implement the method according to any one of claims 1 to 10.

Images