CN118396169A

CN118396169A - Data processing method, system, device and storage medium of intelligent manufacturing factory

Info

Publication number: CN118396169A
Application number: CN202410554191.XA
Authority: CN
Inventors: 张心诚; 马运涛; 劳建音
Original assignee: Shenzhen E Plus Technology Co ltd
Current assignee: Shenzhen E Plus Technology Co ltd
Priority date: 2024-05-07
Filing date: 2024-05-07
Publication date: 2024-07-26

Abstract

The invention relates to the technical field of intelligent manufacturing and industrial automation, in particular to a data processing method, a system, a device and a storage medium of an intelligent manufacturing factory, which integrate Internet of things equipment, a machine learning technology and a digital twin model to optimize the production flow of the intelligent manufacturing factory; by collecting and analyzing critical operational data on the production line in real time, the system is able to generate accurate predictive models based on historical and real time data; the models help to simulate the risk-free production flow on the digital twin platform, identify potential risks and propose process optimization suggestions; in addition, the system combines production data with market data, predicts market demands through machine learning analysis, and supports the establishment of a data-driven production adjustment strategy; through the measures, the invention obviously improves the production efficiency, reduces the operation cost, improves the product quality and the market response speed, and enhances the competitiveness of enterprises.

Description

Data processing method, system, device and storage medium of intelligent manufacturing factory

Technical Field

The present invention relates to the field of intelligent manufacturing and industrial automation technologies, and in particular, to a data processing method, system, device and storage medium for an intelligent manufacturing factory.

Background

With the rapid development of intelligent manufacturing plants, efficient data processing systems are one of the core technologies supporting their operation. Intelligent manufacturing plants rely on accurate data analysis to optimize the configuration of the production line, improve resource utilization, and reduce energy consumption and carbon emissions. However, one major challenge faced by intelligent plants is how to achieve efficient processing of the large amounts of multidimensional data collected.

Problems with current intelligent manufacturing plants in terms of data processing mainly include deficiencies in data integration and real-time data processing. Due to the diversity of production equipment and systems, the data is highly heterogeneous and the data quality is ragged, which makes it difficult to extract valuable information from these data.

The prior art (Chinese patent publication No. CN 116862299B) has defects in dealing with the problems, such as incapability of reflecting the change of the state of a production line in real time, so that the response to abnormal production is not timely enough, and the production efficiency and the product quality are affected. The limited ability to process dynamic data and predict in real time often relies on static models, which limit their utility in rapidly changing production environments. Existing systems often fail to efficiently integrate data from disparate sources, and lack the ability to monitor and manage data quality in real time, which makes data driven decision support systems limited in performance.

Disclosure of Invention

Aiming at the problems existing in the prior art, the invention provides a data processing method, a system, a device and a storage medium of an intelligent manufacturing factory, which are used for collecting key data on a production line in real time by utilizing equipment and sensors of the Internet of things, such as equipment state, production efficiency, raw material consumption and other information; after cleaning and standardization, the data are used for training and updating a machine learning model so as to predict and adapt to production demand changes in real time; meanwhile, a digital twin model is utilized to carry out risk analysis and process optimization, and an optimal production configuration is found through a simulation experiment; finally, combining with market analysis results, formulating and adjusting production strategies to achieve optimal allocation of resources and production adjustment; the core effects include improving production efficiency, reducing cost, improving product quality and enhancing market adaptability.

The first aspect of the present invention provides a data processing method for an intelligent manufacturing plant, comprising the steps of:

The method comprises the steps of collecting key operation data of a production line in real time through Internet of things equipment and sensors, wherein the key operation data comprise equipment state data, production efficiency data and raw material consumption data; cleaning, normalizing and integrating the collected data to generate an integrated production data set;

Establishing an initial machine learning model based on a historical data part of the integrated production data set, continuously inputting real-time data in the integrated production data set into the self-adaptive learning system, updating and adjusting the prediction model in real time, and generating an updated prediction model and a real-time production report;

Creating a digital twin model of the production line by using the updated prediction model and the real-time production report, running simulation on the digital twin model, performing risk analysis and process optimization, and generating a simulation optimization result, wherein the simulation optimization result comprises potential risk assessment and improvement suggestions;

Integrating the simulation optimization result with market data, analyzing by applying a neural network or a machine learning model, predicting market and production requirements, generating a comprehensive analysis report, and providing resource allocation and production adjustment suggestions based on data driving;

adjusting a production strategy according to the suggestion of the comprehensive analysis report, wherein the production strategy comprises production line configuration, human resources and raw material purchasing plans; carrying out adjustment and collecting adjusted production data, evaluating the actual effect of adjustment measures, and generating a production effect evaluation report; and feeding back improvement measures according to the production effect evaluation report, and continuously optimizing the self-adaptive learning model and the production flow.

Preferably, the device status data includes: machine run time, temperature, vibration level, production efficiency data include: yield rate, downtime, raw material consumption data including: raw material type and amount used.

Preferably, the adaptive learning system is configured to adapt to and reflect immediate changes in the production process, wherein the machine learning model uses real-time data in the production dataset for self-adjustment and optimization, and updates the model weights and parameters according to new data without reloading the entire production dataset to generate an updated prediction model and a real-time production report for reflecting new production environment changes in real time, and adjusts its predictions in time to adapt to the latest conditions of the production line.

Preferably, a parallel virtual model is created by digital twin technology using an updated predictive model and a real-time production report, and the digital twin model is obtained, and the updated predictive model includes: the real-time production report comprises: immediate feedback of equipment performance and production quality.

Preferably, after the digital twin model is created and initialized, the digital twin model is used for running production simulation in a virtual environment and testing a new optimization strategy, identifying potential production bottlenecks, equipment failure points and other potential risk factors causing production efficiency degradation or quality problems, completing risk analysis and process optimization, and generating simulation optimization results.

Preferably, the simulation optimization result and the market data are integrated into a unified analysis framework, a neural network or a machine learning model is applied to analyze the integrated data, potential modes and correlations are identified based on historical and current data and considering potential market variation and the influence of production strategies, and future market and production demands are predicted; wherein the market data comprises: consumer behavior trends, market demand variations, and raw material price variations.

Preferably, the optimization of the adaptive learning model is completed by adjusting parameters, algorithms or data input in the adaptive learning model to better predict and cope with actual conditions encountered in production; the production process is optimized by changing production steps, optimizing resource allocation or introducing a new operation protocol so as to improve production efficiency and product quality.

A second aspect of the present invention provides a data processing system for an intelligent manufacturing plant, comprising:

The data acquisition unit is configured with the internet of things equipment and the sensor and is used for collecting key operation data of the production line in real time, wherein the key operation data comprise: equipment status data, production efficiency data, and raw material consumption data;

The data processing unit is used for cleaning, normalizing and integrating the collected data to generate an integrated production data set;

A prediction model generating unit configured with a machine learning algorithm for establishing an initial machine learning model based on the historical data portion of the integrated production data set, and continuously inputting real-time data in the integrated production data set into the adaptive learning system to update and adjust the prediction model in real time;

a report generation unit for generating a real-time production report based on the updated prediction model and the real-time production data;

The digital twin model unit is used for creating a digital twin model of the production line by using the updated prediction model and the real-time production report, running simulation on the model, carrying out risk analysis and process optimization, and generating simulation optimization results, including potential risk assessment and improvement suggestions;

the market analysis unit is configured with a neural network or other machine learning model and is used for integrating and analyzing the simulation optimization result and market data, predicting market and production requirements, generating a comprehensive analysis report and providing data-driven resource allocation and production adjustment suggestions;

The production strategy adjusting unit is used for adjusting the production strategy according to the suggestion of the comprehensive analysis report, including production line configuration, manpower resources and raw material purchasing plans, and implementing adjustment;

The effect evaluation unit is used for collecting the adjusted production data, evaluating the actual effect of the adjustment measures and generating a production effect evaluation report;

And the feedback optimization unit is used for reporting feedback improvement measures according to the production effect evaluation and continuously optimizing the self-adaptive learning model and the production flow.

A third aspect of the present invention provides a data processing apparatus of an intelligent manufacturing plant, comprising:

The data acquisition device is provided with Internet of things equipment and a sensor and is used for collecting key operation data of a production line in real time, wherein the key operation data comprise: equipment status data, production efficiency data, and raw material consumption data;

the data preprocessing device is used for cleaning and standardizing the acquired data and integrating the data to generate an integrated production data set;

The machine learning model generating device comprises a processor and a memory, wherein the memory stores a program which guides the processor to establish an initial machine learning model by using a historical data part based on an integrated production data set, continuously inputs real-time data in the integrated production data set into an adaptive learning system, and updates and adjusts a prediction model in real time;

Report generating means for generating a real-time production report based on the updated prediction model and the real-time production data;

The digital twin simulation device is used for creating a digital twin model of the production line by using the updated prediction model and the real-time production report, running simulation on the model, carrying out risk analysis and process optimization, and generating a simulation optimization result, and comprises the following steps: potential risk assessment and improvement advice;

The market analysis device comprises at least one neural network or machine learning model, is used for integrating and analyzing simulation optimization results and market data, predicting market and production requirements, generating comprehensive analysis reports and providing data-driven resource allocation and production adjustment suggestions;

The production strategy adjusting device is used for adjusting the production strategy according to the suggestion of the comprehensive analysis report, and comprises the steps of adjusting the production line configuration, manpower resources and raw material purchasing plans, and monitoring the implementation of adjustment;

the effect evaluation device is used for collecting the adjusted production data, evaluating the actual effect of the adjustment measures and generating a production effect evaluation report;

and the feedback optimization device is used for reporting feedback improvement measures according to the production effect evaluation and continuously optimizing the self-adaptive learning model and the production flow.

A fourth aspect of the present invention provides a storage medium having stored thereon a computer program which when executed by a processor performs the steps of a data processing method of the intelligent manufacturing plant.

Compared with the prior art, the invention has the advantages that:

through the self-adaptive learning system, the invention realizes continuous input of real-time data and immediate update of the model, thereby ensuring that the prediction model is always synchronous with the latest production state, and improving the response speed and accuracy to the change of the production line.

By using the digital twin technology to create an accurate production line virtual model, the invention can simulate and test different production schemes in a risk-free environment, and effectively perform risk analysis and process optimization.

By integrating the simulation optimization result and market data and applying a neural network and a machine learning model to carry out deep analysis, the invention predicts future market and production demands and provides data-driven resource allocation and production adjustment suggestions for intelligent manufacturing, thereby improving production efficiency and product quality and ensuring the competitiveness of intelligent factories in a dynamic market environment.

Drawings

FIG. 1 is a schematic flow chart of the method of the present invention;

FIG. 2 is a block diagram of the system of the present invention;

Fig. 3 is a block diagram of the structure of the device of the present invention.

Detailed Description

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is only exemplary and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the present disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. In addition, in the following description, descriptions of well-known structures and techniques are omitted so as not to unnecessarily obscure the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and/or the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It should be noted that the terms used herein should be construed to have meanings consistent with the context of the present specification and should not be construed in an idealized or overly formal manner.

Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a convention should be interpreted in accordance with the meaning of one of skill in the art having generally understood the convention (e.g., "a system having at least one of A, B and C" would include, but not be limited to, systems having a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a formulation similar to at least one of "A, B or C, etc." is used, in general such a formulation should be interpreted in accordance with the ordinary understanding of one skilled in the art (e.g. "a system with at least one of A, B or C" would include but not be limited to systems with a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).

Some of the block diagrams and/or flowchart illustrations are shown in the figures. It will be understood that some blocks of the block diagrams and/or flowchart illustrations, or combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the instructions, when executed by the processor, create means for implementing the functions/acts specified in the block diagrams and/or flowchart. The techniques of this disclosure may be implemented in hardware and/or software (including firmware, microcode, etc.). Additionally, the techniques of this disclosure may take the form of a computer program product on a computer-readable storage medium having instructions stored thereon, the computer program product being for use by or in connection with an instruction execution system.

As shown in fig. 1, the data processing method of the intelligent manufacturing plant includes the following steps:

in intelligent manufacturing systems, the real-time collection of data is accomplished through a series of internet of things (IoT) devices and sensors. These devices are typically deployed at critical production line locations, such as machine equipment, production workstations, conveyor systems, etc., to monitor and record various operating parameters.

The equipment status data includes machine run time, temperature, vibration level, etc. For example, temperature sensors may monitor the temperature of the heating device, while vibration sensors may detect abnormal vibrations of machine components, which data is critical for preventive maintenance.

The production efficiency data includes: throughput rate, downtime, and production cycle time. For example, a line sensor may record the number of units produced per hour, as well as any events that lead to production pauses.

Raw material consumption data includes: the type of raw material, the amount used and the consumption rate. For example, a weight sensor is used to monitor the weight of raw material entering the production line, ensuring that the material consumption is synchronized with the production demand.

The data are transmitted to a central database or a cloud platform in real time through the Internet of things equipment so as to establish a real-time data stream.

Data cleansing and normalization is the process of converting raw data into high quality data that can be used for further analysis and modeling.

Data cleansing involves removing erroneous data, repeating records, correcting format inconsistencies, and filling in missing values. For example, if a sensor records unreasonably high temperature values due to a fault, such data may be identified and deleted or corrected during the cleaning process.

Data normalization ensures that data from different sources has a consistent format and scale, making it suitable for processing by machine learning models. For example, converting temperature from degrees Fahrenheit to degrees Celsius, or converting all production efficiency data to a standard measure of production per hour.

Integrating data involves merging multiple data streams into a unified data set, which includes a combination of multiple data types, such as time series data, transaction data, and log data. In smart manufacturing, the integrated production dataset includes time synchronized data obtained from the various sensors, as well as production lot information and product quality records associated with these data.

The accurate data of the invention makes decisions more based on actual production conditions rather than estimation or experience; the production flow is timely adjusted through a real-time monitoring and predicting model, so that the downtime and resource waste are reduced; accurate raw material consumption data can help to manage inventory more effectively, reducing the risk of overstock or under-supply.

The running time is used for recording the total running time from the starting up to the shutting down of the equipment. The monitoring of the run time helps to determine the utilization of the equipment and predict maintenance requirements. For example, certain components may require replacement or maintenance after a certain number of hours of operation.

The temperature sensor is used for monitoring the temperature of key parts of the equipment so as to prevent equipment faults caused by overheating. For example, overheating indicates insufficient lubrication or mechanical failure.

Vibration level, vibration sensors can detect abnormal vibrations in machine operation, which is often an early warning of mechanical failure. Abnormal vibrations are caused by bearing damage, unbalance, or loose parts.

The yield rate represents the number of units of production completed in a particular time. This is a key indicator of line efficiency, low throughput rate indicates equipment failure or bottlenecks in the production flow.

The downtime records the time at which the production line was not scheduled to stop due to a malfunction or maintenance. Minimizing downtime is critical to improving production efficiency and reducing costs.

Monitoring the type of raw material and its usage is critical to cost control and inventory management. Accurate tracking of consumption of each raw material can help ensure timeliness and cost effectiveness of the material supply. For example, if the rate of consumption of a certain material exceeds an expected rate, the system may automatically trigger a reorder flow to avoid production delays.

The real-time data monitoring and analysis of the present invention allows problems to be discovered and resolved before affecting production; by analyzing the equipment state data, a maintenance plan based on conditions is formulated instead of fixed period maintenance, so that unnecessary maintenance cost is reduced and unexpected shutdown caused by equipment failure is avoided; accurate raw material consumption data can optimize inventory levels, reducing excessive inventory and stock-out risks; the integrated data view provides a panorama of the production line, helping the management layer to make better strategic decisions.

the historical portion of the integrated production dataset is used, including historical equipment status data, production efficiency data, and raw material consumption data. These data reflect past production and results, providing the necessary inputs for modeling. An appropriate machine learning algorithm (e.g., regression analysis, decision tree, support vector machine, or neural network) is selected to learn patterns and relationships in the data. For example, historical data is used to train a classification model that predicts equipment failure, which predicts whether it is about to fail based on the temperature and vibration level of the equipment.

In a digital twin environment, common machine learning algorithms include:

Neural networks, because of their powerful nonlinear modeling capabilities, are suitable for processing complex manufacturing data and pattern recognition tasks. The neural network may adjust the weights in real time by a back propagation algorithm to accommodate the new data input.

Decision trees, suitable for classification tasks, can learn new data points incrementally, adapt to data changes by creating new tree nodes or adjusting existing branching conditions.

The newly collected real-time data in the integrated production dataset is continuously input into the trained machine learning model. This ensures that the model is able to receive up-to-date production line information, which is critical to the dynamic production environment. The system can automatically adjust and optimize model parameters according to the newly input real-time data. This adaptive capability allows the model to learn and adapt to actual changes in production conditions rather than relying solely on historical data. By continuously inputting new real-time data and feedback, the model gradually optimizes its prediction accuracy. For example, if the model predicts a device failure that deviates from the actual failure onset, the model adjusts its internal parameters to more accurately predict future failures.

The data provided by the real-time production report, such as equipment performance and production quality feedback, is directly input into the digital twin model. These data are used to:

the prediction model is adjusted, and real-time data can be used to verify the accuracy of model prediction and adjust model parameters to reduce prediction errors.

Simulating future conditions, the digital twin model can simulate different production strategies by using the latest data, and predict potential influences on production efficiency and product quality.

The output of the digital twin model may support decision making such as adjusting production speed, altering raw material usage policies, or scheduling equipment maintenance. Model prediction and risk assessment results can guide the production line to make corresponding adjustments to optimize operation and enhance flexibility of the production line.

The updated model can output a prediction result, such as a future state of the apparatus, an expected change in production efficiency, or a prediction of raw material consumption, based on the latest data. The system automatically generates a real-time production report containing key performance indicators and predicted results. These reports provide immediate data support for production management, helping the management layer to make quick and informative decisions.

The invention can reflect and adapt to the change of the production line in real time through the continuously updated prediction model, thereby optimizing the production flow and reducing the downtime; the self-adaptive prediction model can more accurately predict equipment faults and maintenance requirements, and prevent maintenance rather than post repair is realized, so that the risk and cost of unexpected faults are reduced; the generated real-time production report provides real-time data analysis for the management layer, and supports faster and more data-driven decision processes.

Real-time data, such as plant operating conditions, production rates, and raw material usage collected from sensors and monitoring equipment, is continuously input into a pre-trained machine learning model. These data reflect current production conditions and equipment performance.

Real-time production reporting is a critical output in intelligent manufacturing systems that provides the necessary information for production management to make quick and accurate decisions. The real-time production report includes:

The device status overview includes data of the current operating status, temperature, vibration level, etc. of all critical devices. This helps in identifying in time the equipment that needs maintenance or fails.

And the production efficiency index displays the output rate, the comparison of the plan and the actual output, the downtime and the reason analysis thereof. These data help the management layer monitor production efficiency and respond quickly to any production delays.

Raw material consumption, reporting the usage amount of each raw material, and comparing with a preset consumption plan to monitor whether the raw material usage is efficient or not and whether waste exists or not.

Quality control data, including results of product quality detection during production, such as yield and defective product analysis, are critical to quickly identify quality problems and to adjust.

Prediction and warning, impending problems predicted based on the latest models, such as equipment failures, production bottlenecks, etc., and suggested precautions.

In conventional machine learning practice, models are typically trained and tested on static data sets. In the self-adaptive learning system, the model dynamically updates the parameters by using the real-time data stream, and the whole model does not need to be retrained. This is achieved by online learning techniques such as incremental learning or updating weights using gradient descent methods. For example, if a model is used to predict equipment failure, it will adjust its prediction threshold and sensitivity based on temperature and vibration data collected from the equipment in real time.

In an ideal case, the adaptive learning model should be continuously updated, i.e. the model receives new data in real time and adjusts its parameters instantaneously. The method is suitable for a highly dynamic production environment, and can reflect the latest production conditions and data changes in real time.

In some cases, the model may also be updated at fixed intervals (e.g., once per hour or per shift). This applies to production processes that vary relatively smoothly or when real-time processing resources are limited.

The triggering conditions include:

(1) Data bias triggers:

When there is a significant deviation of the real-time data from the model predictions, an update is triggered. For example, if the model predicted device state differs significantly from the actual detected device state, the system will readjust the model to more accurately reflect the actual condition of the device.

(2) Performance index reduction triggering:

Model updates may also be triggered if key performance indicators (e.g., equipment efficiency, product quality) fall below a certain threshold. This indicates that the existing model cannot accurately capture the characteristics of the current production process.

By setting reasonable updating frequency and triggering conditions, the model can be ensured to always maintain the optimal state, and stable and accurate decision support is provided for intelligent manufacturing. This adaptive mechanism helps to improve production efficiency, reduce downtime, and ultimately achieve cost savings and production improvements.

The system can adjust internal parameters, such as weights, of the model in real time according to the newly received data. Such updating ensures that the output of the model continuously reflects the latest production state, enhancing the accuracy and relevance of the predictions.

The updated model can map the current production environment more accurately, predicting future trends and potential problems. Meanwhile, a real-time production report automatically generated by the system provides an immediate data view for a management layer and supports quick decision-making.

In the invention, the self-adaptive learning system allows the production line to quickly respond to environmental changes, such as adjusting the raw material supply quantity to respond to the change of the raw material consumption speed, or adjusting the production speed to adapt to the fluctuation of the equipment performance; by continuously learning the latest production data, the model can more accurately predict potential problems such as equipment faults, production bottlenecks and the like, so that measures are taken in advance to avoid production interruption and reduce maintenance cost; the real-time production report provides insight for the management layer based on the latest data, so that decisions are driven by more data and react faster, and the efficiency and response speed of the whole production flow are enhanced.

The digital twin model is based on detailed replication of the physical production line, including all relevant machine equipment, transmission systems and operational flows. This model is built using data collected from the physical production line (e.g., equipment status data, production efficiency data). The updated prediction model provides data support for digital twinning, and ensures that the virtual environment and the actual production environment are kept synchronous.

The real-time production report provides continuous data input, such as the instantaneous operating state of the machine and production data, so that the digital twin model can reflect the latest conditions of the production line in real time.

The simulation is run on a digital twin model simulating various production scenarios and operating conditions such as equipment speed variation, raw material replacement or production flow adjustments. This helps identify potential production bottlenecks or failure points. By simulating different production situations, the risk of production interruption or efficiency degradation is analyzed. For example, if the simulation shows frequent malfunctions at a particular plant speed, this indicates that the speed setting is not suitable for the current production conditions. Based on the simulation results, the system may generate improvement recommendations, such as adjusting tact, improving equipment maintenance planning, or optimizing raw material usage.

In the present invention, the digital twinning model allows enterprises to test various operational strategies without affecting actual production, thereby identifying and mitigating potential production risks. This precaution reduces accidental downtime and loss of production. Through real-time monitoring and simulation, the management layer can further understand each link of the production flow, and finer production control and decision support are realized. According to the improvement proposal of the simulation optimization result, effective process optimization measures are implemented, so that the production efficiency can be remarkably improved, and the extra cost caused by low efficiency is reduced.

The digital twin technology provides a risk-free experimental platform on which enterprises can test new ideas and strategies to continuously optimize the production process. The application of the technology not only improves the flexibility and response capability of production, but also enhances the adaptability of enterprises to market changes.

Digital twinning is a high-level technique that mirrors a physical production line by creating a virtual model. The core of this technique is the ability to simulate and predict various conditions in the production process in real time, thereby making more efficient decisions and optimizations.

The digital twin model is constructed by integrating real-time production data, historical data and predictive models. The virtual model not only reflects the current production line state, but also simulates future running conditions. For example, if a production line includes numerically controlled machine tools, assembly robots, and conveyor belts, digital twinning would rebuild these devices in a virtual environment and update their status and performance in real time.

The model may be continuously adjusted based on data collected in real time, such as adjusting the time to predict equipment failure or predicting the nature of the production bottleneck based on machine learning algorithms. The model automatically updates its parameters based on new data collected from sensors, etc. Including immediate data of plant performance and production quality feedback such as plant operating efficiency, production quality, etc. These reports provide an immediate view of the data for model analysis and decision support.

Online learning is a learning technique in dynamic data streams where models are continually receiving new data and updated on the fly. This avoids the need to reload and train the entire data set, enabling the model to adapt quickly to new situations.

Migration learning, in some cases, models that have been trained on similar tasks can quickly adapt to new production environments through migration learning, with only small amounts of parameters being trimmed.

Simulations are run on the virtual model to test different production configurations or coping strategies in real time, such as adjusting production speed or altering raw material supply chains, to see how these variations affect production efficiency and product quality. The simulation can reveal potential risks and bottlenecks, and provide scientific basis for adjustment in the actual production process. For example, if the simulation shows that increasing the operating speed of a machine results in an increase in failure rate, the management layer may choose to reduce the speed appropriately to avoid frequent maintenance.

The invention can help the factory to reduce unexpected downtime and optimize maintenance plan by accurately predicting equipment faults and production problems, thereby reducing maintenance cost and improving equipment use efficiency. Real-time simulation and prediction enables the plant to better plan production activities, adjusting production strategies to accommodate market demand changes and raw material supply conditions. In addition, by monitoring the production quality in real time, measures can be taken rapidly to correct the deviation, so that the product quality is ensured. The real-time data and predictive reports provided provide decision support for the management layer, helping them to make scientific decisions based on the latest production line data and predicted results.

In the context of intelligent manufacturing, a digital twin model is first constructed by integrating and analyzing real-time and historical data of a plant, including equipment status data, production efficiency data, and raw material consumption data. This integration process involves the cleaning, standardization of data, and real-time updating of data, ensuring that the digital twin model reflects the latest state of the physical environment.

The initialization phase, the digital twin model is configured in a virtual environment, which covers the setting of model parameters, the programming of production logic and the calibration synchronized with the real production line. For example, if a temperature sensor on one production line shows frequent abnormal readings during the past week, this information will be used to adjust the behavior prediction logic of the corresponding device in the digital twin model.

After the digital twin model is initialized, the model is used to run various production simulations. These simulations are performed by software tools without actual physical risk, enabling simulation of different production scenarios, such as changing production line speeds, adjusting raw material input ratios, etc. These simulations help identify potential bottlenecks and failure points in the production process, for example, by simulating increases in product line speed, a significant increase in failure rate of a critical machine may be observed, indicating that the apparatus is a production bottleneck at high loads.

Through the simulation, the digital twin model can identify potential risk factors causing production efficiency degradation or quality problems. For example, if the simulation shows that the product defect rate rises when a particular lot of raw material is used, this indicates that there is a problem with raw material supply. In addition, the model may test new optimization strategies such as adjusting raw material mixing ratios or changing the order of operation of certain workstations to see the impact of these changes on production efficiency and product quality.

Different configuration schemes of production line operation can be simulated by using the digital twin model to evaluate the influence of various process changes on production efficiency. For example, by adjusting the order of the machines or the time intervals of operation on the production line in a simulation, an optimal production line layout can be found to reduce production delays and raw material wastage. In particular, if a delay in a particular process is found to affect a subsequent process, a model may be used to test the effect of advancing the process or adding parallel workstations.

By means of digital twinning technology, different settings of the parameters of the device can be tested without affecting the actual production. For example, for injection molding machines, adjustments to temperature, pressure, or injection rate can be simulated to determine optimal combinations of parameters that will improve product consistency and quality while reducing energy consumption.

The digital twin model can be used to evaluate the potential impact of replacement suppliers or raw materials on product quality and production costs. By simulating the use of raw materials of different quality or cost, variations in product yield and possible cost savings can be predicted, helping the decision maker to select the optimal supply chain strategy.

By running the simulation in a virtual environment, the digital twin model is not only able to predict and identify problems in the production process, but also to test the solution strategy prior to actual application. This reduces the cost and downtime in actual testing, improving the reliability and efficiency of the production line. The finally generated simulation optimization result provides detailed data support, and provides scientific basis for a decision-making layer of a manufacturing factory, so that more effective production strategies and risk management measures are formulated.

Before production simulation, it is first necessary to ensure that the digital twin model accurately reflects all key variables and parameters of the actual production environment. This includes collecting data from the production line in real time, such as equipment operation data, environmental conditions, and operator inputs, and importing historical production data from ERP and MES systems.

When the simulation is performed, specific test scenarios such as production peak period, equipment failure simulation, raw material shortage scenario and the like need to be set to observe the performance of the model under various pressure tests.

To ensure accuracy of the simulation, the model needs to be calibrated periodically to match the actual production data, including adjusting input parameters of the model, updating algorithms, or introducing new data points. Furthermore, advanced data analysis and machine learning techniques may be employed to improve the accuracy of predictions, such as automatically adjusting model parameters to accommodate new production trends using machine learning algorithms.

Reliability is ensured by the repeatability and stability of the simulation. The same simulation scene is repeatedly operated, and the consistency of the comparison results can be verified, so that the stability and the reliability of the model can be verified.

Simulation results provide insight into potential problems and opportunities for improvement in the production process. By analyzing these results, specific links that need to be optimized, such as adjusting production cadence, improving quality control flow, or optimizing raw material usage, can be identified.

After the improvement measures are implemented, the digital twin model is continuously used for monitoring the improvement effect, and iterative optimization is performed. This continuous feedback loop ensures that the production process is continually adapted to new operating conditions and market demands.

The principles and effects of digital twin models in intelligent manufacturing, particularly important roles in production optimization and risk management, can be more fully understood from such detailed description and illustration.

The digital twin technology is applied to intelligent manufacturing factories, and provides a strong technical support for improving production efficiency, reducing operation risks and improving product quality through highly accurate simulation and optimization analysis. The technical means effectively realize prediction, prevention and optimization, and remarkably improve the overall performance of intelligent manufacturing.

Using the historical data, a machine learning model may be constructed to identify abnormal patterns of the device. Common techniques include Support Vector Machines (SVMs), random forests, and neural networks, which are capable of learning from operational data of the device (e.g., temperature, vibration frequency, energy consumption, etc.) and predicting potential faults or time at which the fault occurred. For example, by analyzing vibration data, a trained model can identify early signs of bearing failure.

Using process simulation software (e.g., arena or Simul 8), the effects of different variables in the production process are simulated, thereby identifying production bottlenecks. By adjusting input parameters (e.g., raw material supply speed, machine speed, etc.), different production scenarios are simulated, thereby identifying which factors most lead to production delays or quality problems.

The dynamic behavior of the whole production system is simulated by applying a system dynamics method, and key variables affecting the production efficiency and quality are identified. By building a causal graph and feedback loop reflecting the production flow, the long-term impact of policy changes or external environmental changes on the production system can be predicted.

Simulation optimization results typically include quantitative data regarding the effects of various production parameter variations. These data may be presented to the decision maker in the form of a chart or dashboard through a data visualization tool (e.g., tableau or Power BI) to help them understand the potential impact of different decision schemes and select the optimal scheme.

According to the simulation result, the configuration of the production line can be adjusted, such as the production line is rearranged, the material flow is optimized, or the equipment with lower efficiency is replaced. For example, if a simulation shows that a certain production segment often becomes a bottleneck, the workstation of that segment may be increased or automation techniques may be introduced to increase processing capacity.

After any improvement is implemented, it is important to continuously monitor the actual effect of these changes. This can be done by comparing the actual production data with the simulated predictions. If the actual effect does not reach the expected value, the model needs to be adjusted or other optimization measures are explored.

With the initially implemented feedback, a new round of simulation and analysis continues using the digital twin model. The iterative method helps enterprises to refine the production flow step by step and continuously adapt to changing markets and technical conditions.

Through such detailed analysis and practical application description, the role of digital twin technology in the intelligent manufacturing environment is fully revealed, and particularly, the important value in optimizing the production process and improving the decision quality is realized. The application of these techniques ensures that the manufacturing enterprise is able to maintain flexibility and competitiveness in a highly competitive market place.

Integration of the simulation optimization results from the digital twin model with external market data is a critical step. This includes integrating internal production data (e.g., plant efficiency, failure rate, yield quality) with external data (e.g., market demand changes, raw material price fluctuations, competitor actions). Data integration is typically accomplished through a data warehouse or data lake environment, ensuring consistency and queriability of data. The integration process of data is automated using ETL (extract, transform, load) tools, such as APACHE NIFI or Talend. These tools support extracting data from a variety of data sources, converting the data to meet analysis needs, and loading onto an analysis platform.

The integrated data is subjected to deep analysis by applying a neural network or a machine learning model. These models are able to identify complex patterns and trends in the data, predicting changes in market and production demand. For example, long-term memory network (LSTM) models are used to analyze time-series data to predict future market demand, or decision tree models are used to evaluate the impact of different production configurations on product quality. Neural networks are particularly useful for processing large-scale data sets, from which valuable features can be automatically learned and extracted by deep learning techniques.

Based on the model analysis results, a comprehensive analysis report is generated detailing the prediction results and their impact on the current production strategy. Reporting helps decision makers understand complex data and model predictions through data visualization. Using business intelligence tools, such as Power BI or Tableau, an interactive dashboard is created that exposes Key Performance Indicators (KPIs), trend analysis, and predictive results. These tools support advanced analysis functions, how to optimize resource allocation and production adjustments in a data driven manner.

The invention uses integrated data and advanced analytical models, and the intelligent manufacturing factory can more accurately predict market demand and production demand. For example, by analyzing historical sales data and market trends, the model predicts that a certain product line will experience demand peaks in the coming months, thereby adjusting the production plan in advance. The integrated analysis of digital twinning technology and market data enables manufacturing enterprises to quickly adapt to market changes. For example, if the model analysis shows that raw material costs are rising, the enterprise may adjust purchasing strategies in time to find lower cost suppliers or alternative materials to maintain profit margins. The data driven insight provided by the analysis-by-synthesis report can be used to optimize resource allocation and production adjustments. For example, if predictive analysis indicates that the production efficiency of a certain product is low, the enterprise decides to move more resources to a more efficient or profitable product line to increase overall production efficiency and profitability.

The invention integrates the simulation optimization result of the digital twin model with market data, and applies advanced machine learning technology to analyze, thereby providing a powerful tool for intelligent manufacturing factories, optimizing production decision and responding market change by scientific and systematic methods, and improving competitiveness and market adaptability.

Integrating the simulated optimization results from the digital twin model with the market data into a unified analysis framework is a key step that involves normalizing and synchronizing structured data (e.g., production data, raw material costs) and unstructured data (e.g., market trend reports, consumer behavior analysis). With data integration platforms such as APACHE KAFKA or APACHE NIFI, these tools support high throughput data stream processing that can process and forward data to an analysis system in real time. Meanwhile, a data lake architecture such as Amazon S3 or Hadoop HDFS is used for storing a large amount of heterogeneous data, so that large-scale data analysis is convenient.

In a unified analysis framework, neural networks and other machine learning models are used to analyze the integrated data. These models can learn from historical and current data and predict changes in market and production needs. For example, a Convolutional Neural Network (CNN) may be trained to identify patterns of raw material price fluctuations, while long and short term memory networks (LSTM) are adapted to analyze time series data to predict market demand trends.

When the model is predicted, not only historical data is considered, but also potential influences of market dynamic changes and internal production strategy adjustment on the future are analyzed. Scene analysis and sensitivity analysis tools are used to evaluate the impact of different market-changing scenarios (e.g., economic decay, consumer preference changes) on production needs. This analysis helps predict production requirements and resource allocation requirements under different conditions.

By integrating the simulation results and the market data, the neural network and the machine learning model can provide more accurate market and production demand prediction. For example, if model analysis finds that an increase in raw material price is associated with a decrease in certain market demands, the production plan may be adjusted accordingly to reduce costs and inventory backlog.

Integrated analytics enable enterprises to quickly adapt to market changes. For example, if the model predicts that demand for a certain type of product will increase over a period of time in the future, the enterprise may adjust the production line in advance to increase the throughput of the product to capture market opportunities.

The analysis-by-synthesis report is based on data-driven insights that can be used to optimize resource allocation and production strategies. This includes distribution of productivity, timing and scaling of raw material purchases, and possible equipment upgrades or replacements to ensure production efficiency is maximized and cost is reduced.

The simulation optimization result of the digital twin model is integrated with market data, and advanced machine learning and neural network technology are applied to analyze, so that not only is the accuracy of prediction improved, but also the response capability of manufacturing enterprises to market changes is enhanced, and the competitive advantage is maintained in the highly competitive market.

Adjusting the production strategy involves modifying the line configuration, optimizing human resource allocation, and rescheduling the raw material procurement plan according to the recommendations of the integrated analysis report. These adjustments are based on data-driven insights with the aim of improving production efficiency, reducing costs, and improving product quality. For example, if the analysis report indicates that the output of a certain production line is lower than expected, it is recommended to add automated equipment or retrain the operator. For raw material purchase, if the report shows that a raw material price is expected to rise, it is recommended to purchase or find an alternative material in advance.

After the adjustments are made, it is important to collect adjusted production data, including but not limited to production speed, product yield, equipment failure rate, etc. These data are used to evaluate the actual effect of the adjustment measures. The operating state of the production line is monitored in real time using internet of things (IoT) devices and sensors, data is automatically collected, and the data is consolidated and analyzed by an industrial internet platform.

Based on the collected data, a production effect evaluation report is generated, which details the effect of the adjustment measures, including changes in production costs, improvement in production speed, improvement in product quality, and the like. Reporting is accomplished through advanced data analysis and visualization techniques, such as using SAP CRYSTAL Reports or Microsoft Power BI, etc. tools to present detailed analysis results and trend graphs.

Based on the production effect evaluation report, the adaptive learning model and the production flow are further adjusted. This includes parameter adjustment of the model, re-optimization of the production flow, etc., ensuring that the production system can continuously adapt to changing markets and technical environments. For example, if the assessment report shows an increase in the failure rate of a device, it may be necessary to adjust the algorithm for predicting failure in the adaptive learning model, or to add more frequent maintenance and inspection to the production flow.

By continuously optimizing the production strategy and flow, the production efficiency can be obviously improved, and the cost caused by equipment failure and unqualified products can be reduced. For example, a more accurate raw material procurement plan may reduce inventory costs and avoid material wastage.

The product quality is continuously improved through fine adjustment and optimization of the production flow. For example, optimized human resource allocation and equipment management may reduce operational errors and equipment failures, thereby improving the yield of the final product.

Continuous data analysis and feedback mechanisms enable enterprises to quickly respond to market changes, such as demand fluctuations or raw material price changes, and maintain the enterprises' leading position in competition.

The adaptive learning model continually receives new input data (e.g., equipment performance data, product quality data, and operating parameters) in the intelligent manufacturing environment and adjusts the model parameters, algorithms, or overall architecture based on such data to more accurately predict and reflect production changes. For example, the weights and parameters of the model may be updated in real-time using machine learning techniques such as online learning algorithms. Assuming that new automated equipment is introduced on the production line, the model learns the behavior pattern of the equipment through real-time data, and adjusts the prediction algorithm to adapt to the new production environment. On a production line, if an increase in product defect rate is detected, the adaptive model may analyze what variables (e.g., temperature fluctuations, material lot variations) are caused and automatically adjust the impact weights of these variables to optimize the quality control model in real time.

Optimizing production steps, resource allocation and operating protocols based on the data and predictions provided by the adaptive learning model. This includes adjusting the operating speed of the production line, reconfiguring the production shift, or introducing a new quality check procedure. The production steps are analyzed and optimized using methods of operations planning and industrial engineering, such as linear programming and queuing theory. These methods can help determine the optimal production cadence and resource allocation to maximize production efficiency. On an electronic assembly line, it is found that a certain inspection step often becomes a bottleneck. Through data analysis, the front-end and rear-end processes are adjusted, the load is balanced, and an automatic visual detection system is introduced to accelerate the detection speed and reduce the dead time.

According to the invention, the accuracy of production prediction is improved by continuously optimizing the self-adaptive learning model, so that the production plan is more accurate, and the resource waste and excessive production are reduced.

Optimizing the production flow and adjusting the resource allocation can directly improve the operation efficiency of the production line and reduce the faults and the downtime. Meanwhile, the consistency and quality standard of the product can be remarkably improved by improving an operation protocol and introducing an efficient quality control step.

The constant optimization of the adaptive learning model and flexible adjustment of the production flow enables the enterprise to quickly respond to market changes, such as raw material supply interruption or sudden demand increases, thereby maintaining competitiveness.

As shown in fig. 2, a data processing system of an intelligent manufacturing plant, comprising:

The data acquisition unit is configured with the internet of things equipment and the sensor and is used for collecting key operation data of the production line in real time, wherein the key operation data comprise: equipment status data, production efficiency data, and raw material consumption data; the unit monitors the production line in real time by utilizing the Internet of things equipment and sensors, and collects key operation data, including equipment state data (such as temperature, pressure, speed and the like), production efficiency data (such as yield, downtime and the like) and raw material consumption data (such as type, quantity and the like). With IoT devices, such as PLCs (programmable logic controllers) and RTUs (remote terminal units), data is captured and sent in real-time to a central system. For example, temperature sensors monitor thermal effects of the welder to prevent equipment failure due to overheating.

The data processing unit is used for cleaning, normalizing and integrating the collected data to generate an integrated production data set; the collected data are cleaned (abnormal values are removed, missing data are filled), standardized (unified measurement units and formatted date and time) and integrated to generate an integrated production data set, and a cleaned and consistent data source is provided for subsequent analysis. Data preprocessing is performed by using data processing software such as Talend or APACHE SPARK to ensure the quality and consistency of the data.

A prediction model generating unit configured with a machine learning algorithm for establishing an initial machine learning model based on the historical data portion of the integrated production data set, and continuously inputting real-time data in the integrated production data set into the adaptive learning system to update and adjust the prediction model in real time; an initial predictive model is built based on historical portions of the integrated production dataset using a machine learning algorithm (e.g., decision tree, random forest, or neural network). The real-time data is then continuously input into the adaptive learning system, and the model is updated and adjusted in real-time to accommodate production variations. For example, LSTM neural network models are used to predict the likely failure time of a device, thereby enabling maintenance ahead of time, reducing unexpected downtime.

The digital twin model unit is used for creating a digital twin model of the production line by using the updated prediction model and the real-time production report, running simulation on the model, carrying out risk analysis and process optimization, and generating simulation optimization results, including potential risk assessment and improvement suggestions; generating a real-time production report based on the updated predictive model and the real-time production data; and simultaneously, using the data to run simulation on the digital twin model for risk analysis and process optimization. The digital twin model simulates a real production line in a virtual environment, allowing risk-free testing of production improvements, such as adjusting production speed or changing sequence of process steps.

The market analysis unit is configured with a neural network or other machine learning model and is used for integrating and analyzing the simulation optimization result and market data, predicting market and production requirements, generating a comprehensive analysis report and providing data-driven resource allocation and production adjustment suggestions; the unit combines the internal simulation optimization result with external market data (such as consumer behavior trend and market demand variation), and performs deep analysis through a neural network or other machine learning models to predict future market and production demands. And analyzing a complex market mode by using a deep learning technology, and providing accurate demand prediction and resource allocation suggestions.

The effect evaluation unit is used for collecting the adjusted production data, evaluating the actual effect of the adjustment measures and generating a production effect evaluation report; and adjusting the production strategy according to the analysis report, such as reconfiguring the production line, adjusting the human resources and purchasing the raw materials. After the adjustment is implemented, new production data is collected, the actual effect of the change is evaluated, and a production effect evaluation report is generated. For example, when a shortage of raw materials is predicted, purchasing strategies are adjusted in advance, and possible influences are evaluated by simulation results to optimize cost and production efficiency.

And the feedback optimization unit is used for reporting feedback improvement measures according to the production effect evaluation and continuously optimizing the self-adaptive learning model and the production flow. Based on the production effect evaluation report, the adaptive learning model and the production flow are continuously optimized. Model parameters are continuously adjusted according to the actual effect so as to ensure the accuracy of the model and the efficiency of the production flow. Continuous improvement and quality management methods such as six sigma are implemented, with continuous iterations and optimization of the production process.

By means of real-time data-driven prediction and adjustment, the system can quickly respond to production and market changes, optimize resource allocation and reduce waste. Through accurate prediction and optimization, unnecessary raw material consumption is reduced, and equipment utilization rate is improved, so that cost is reduced, and product quality is improved. The digital twin model and market analysis help identify potential production and market risks, and make countermeasures in advance to avoid significant losses.

By the intelligent manufacturing factory data processing system, enterprises can realize highly-automatic and intelligent production management, and the competitiveness and market adaptability are remarkably improved.

As shown in fig. 3, a data processing apparatus of an intelligent manufacturing plant includes:

the data acquisition device is provided with Internet of things equipment and a sensor and is used for collecting key operation data of a production line in real time, wherein the key operation data comprise: equipment status data, production efficiency data, and raw material consumption data; the data acquisition device equipped with the Internet of things equipment and the sensors is focused on monitoring and collecting key operation data on the production line in real time, such as equipment state, production efficiency and raw material consumption data. These data are the basis for intelligent manufacturing system decision-making and optimization. For example, vibration sensors are used to monitor machine health, flow meters measure fluid usage in chemical processes, and power sensors evaluate energy utilization efficiency.

The data preprocessing device is used for cleaning and standardizing the acquired data and integrating the data to generate an integrated production data set; the device is responsible for performing the necessary cleaning (removing noise, correcting erroneous readings), normalization (unifying data formats and units of measure) of the acquired data, and integrating the data to form an integrated production dataset. A data processing software platform, such as APACHE KAFKA or APACHE SPARK, is used to process the large number of real-time data streams and prepare the data for further analysis.

The machine learning model generating device comprises a processor and a memory, wherein the memory stores a program which guides the processor to establish an initial machine learning model by using a historical data part based on an integrated production data set, continuously inputs real-time data in the integrated production data set into an adaptive learning system, and updates and adjusts a prediction model in real time; the apparatus runs a stored program through an advanced processor and memory to build a machine learning model using historical data portions of the integrated production dataset and updates the model in real time to reflect the new data. For example, failure prediction is performed using random forest algorithms, or complex pattern recognition and prediction is performed using deep learning models.

Report generating means for generating a real-time production report based on the updated prediction model and the real-time production data; based on the latest predictive model and the real-time production data, a real-time production report is generated regarding the production status, providing key performance indicators and trend analysis. Such as creating a real-time monitoring dashboard using a BI tool, such as a Power BI or Tableau, to enable the management layer to learn about production conditions in real time.

The digital twin simulation device is used for creating a digital twin model of the production line by using the updated prediction model and the real-time production report, running simulation on the model, carrying out risk analysis and process optimization, and generating a simulation optimization result, and comprises the following steps: potential risk assessment and improvement advice; using the latest predictive model and real-time production reports, a digital twin model of the production line is created and simulations run on this model for risk analysis and process optimization. For example, digital twinning techniques are used to test different production configurations in a virtual environment without actually adjusting the physical production line.

The market analysis device comprises at least one neural network or machine learning model, is used for integrating and analyzing simulation optimization results and market data, predicting market and production requirements, generating comprehensive analysis reports and providing data-driven resource allocation and production adjustment suggestions; and integrating the simulation optimization result and the market data, and using a neural network or other machine learning models to conduct deep analysis so as to predict future market and production requirements. If deep learning is utilized to analyze consumer behavior trend, the method is integrated with production data to form comprehensive market demand prediction.

The production strategy adjusting device is used for adjusting the production strategy according to the suggestion of the comprehensive analysis report, and comprises the steps of adjusting the production line configuration, manpower resources and raw material purchasing plans, and monitoring the implementation of adjustment; according to the advice of the comprehensive analysis report, the production strategy including the production line configuration, the human resources and the raw material purchasing strategy is adjusted to cope with the market change predicted by analysis. The line configuration is adjusted to accommodate changing order requirements, such as by an automated system to adjust tact and production lot.

The effect evaluation device is used for collecting the adjusted production data, evaluating the actual effect of the adjustment measures and generating a production effect evaluation report; and collecting the adjusted production data, evaluating the actual effects of the adjustment measures, and generating a production effect evaluation report. The data analysis tool is used to measure the production efficiency, cost and product quality changes before and after adjustment.

And the feedback optimization device is used for reporting feedback improvement measures according to the production effect evaluation and continuously optimizing the self-adaptive learning model and the production flow. And feeding back improvement measures according to the production effect evaluation report, and continuously optimizing the self-adaptive learning model and the production flow. And (3) iterating the parameter adjustment of the model to refine the precision of the model and optimizing the production flow to realize continuous improvement.

Real-time production reports and comprehensive analysis reports provide immediate and accurate data support for a management layer, so that the decision-making speed and quality are improved; the production strategy and the process are quickly adjusted to adapt to the change of market demands, and production delay and surplus inventory are reduced; continuous improvement of production efficiency and stable improvement of product quality are realized through continuous data feedback and model optimization.

A storage medium having stored thereon a computer program which when executed by a processor performs the steps of a data processing method of the intelligent manufacturing plant.

It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flowchart and/or block of the flowchart illustrations and/or block diagrams, and combinations of flowcharts and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory includes volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises an element.

The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A data processing method of an intelligent manufacturing plant, comprising the steps of:

2. The method of claim 1, wherein the equipment status data comprises: machine run time, temperature, vibration level, production efficiency data include: yield rate, downtime, raw material consumption data including: raw material type and amount used.

3. The data processing method of an intelligent manufacturing plant according to claim 1, wherein the adaptive learning system is configured to adapt and reflect immediate changes in the production process, wherein the machine learning model uses real-time data in the production dataset for self-tuning and optimization, and wherein the model is updated with weights and parameters of the new data without reloading the entire production dataset to generate an updated predictive model and a real-time production report that is configured to reflect new production environment changes in real-time, and adjust its predictions in time to accommodate the latest conditions of the production line.

4. The method of claim 1, wherein the digital twin model is obtained by creating a parallel virtual model using the updated predictive model and the real-time production report by digital twin technology, the updated predictive model comprising: the real-time production report comprises: immediate feedback of equipment performance and production quality.

5. The method of claim 1, wherein after the digital twin model is created and initialized, the digital twin model is used to run production simulation in a virtual environment and test new optimization strategies, identify potential production bottlenecks, equipment failure points, and other potential risk factors that lead to production efficiency degradation or quality problems, complete risk analysis and process optimization, and generate simulation optimization results.

6. The method of claim 1, wherein the simulation optimization results are integrated with market data into a unified analysis framework, the integrated data is analyzed using neural networks or machine learning models, the integrated data is analyzed based on historical and current data, and potential patterns and associations are identified and future market and production needs are predicted in consideration of potential market variations and the effects of production strategies; wherein the market data comprises: consumer behavior trends, market demand variations, and raw material price variations.

7. The data processing method of an intelligent manufacturing plant according to claim 1, wherein the optimization of the adaptive learning model is accomplished by adjusting parameters, algorithms or data inputs in the adaptive learning model to better predict and cope with actual conditions encountered in production; the production process is optimized by changing production steps, optimizing resource allocation or introducing a new operation protocol so as to improve production efficiency and product quality.

8. A data processing system for an intelligent manufacturing plant, comprising:

9. A data processing apparatus for an intelligent manufacturing plant, comprising:

10. A storage medium having stored thereon a computer program, characterized in that the program, when being executed by a processor, realizes the steps of the data processing method of an intelligent manufacturing plant according to any of claims 1-7.