Factory data management system and method based on digital twinning

Abstract

The invention discloses a factory data management system and method based on digital twinning, relates to the technical field of factory management, and solves the problems that in the prior art, equipment operation parameters are collected mostly through a sensor, abnormity is judged based on a static threshold value, but coupling influence of environment variables on faults is not fused, misjudgment is easily generated due to environment fluctuation, and meanwhile, the fault rate is low. The comprehensive optimization of the production efficiency is neglected, so that the monitoring and early warning precision is not high; the method comprises the following steps: constructing a digital twinborn model of factory production equipment; the method comprises the following steps: acquiring simulation data during simulation operation of production equipment and environment data at a corresponding moment; calculating an evaluation coefficient based on the simulation data; based on the environment data, constructing a correction model to correct the fault evaluation coefficient, and obtaining a target fault coefficient; predicting the current simulation data based on a pre-trained fault prediction model to obtain a prediction result; and finally, whether early warning is carried out is judged based on a prediction result. The technical problem is solved.

Description

Technical Field

[0001] The present invention belongs to the field of plant management, and specifically relates to a plant data management system and method based on digital twins. Background Art

[0002] With the evolution of industrial digitalization, factory data management is becoming increasingly complex. Traditional systems, with their shortcomings in data integration, real-time monitoring, and simulation prediction, struggle to meet the demands of efficient modern factory operations.

[0003] In existing technologies, factory equipment management usually relies on manual inspections and isolated data systems. Basic sensors are used to collect equipment operating parameters and judge anomalies based on static thresholds. However, this method has significant defects: First, data fragmentation leads to one-sided equipment status assessment, without integrating the coupling effect of environmental variables on faults, which can easily lead to misjudgment due to environmental fluctuations; second, the evaluation indicators are single and ignore the comprehensive optimization of production efficiency, resulting in low monitoring and early warning accuracy.

[0004] Therefore, the present invention solves the above problems by proposing a plant data management system and method based on digital twins. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a plant data management system and method based on digital twins, which is used to solve the technical problems that most of the prior arts collect equipment operating parameters through sensors and judge anomalies based on static thresholds, but do not integrate the coupling influence of environmental variables on faults, are prone to misjudgment due to environmental fluctuations, and at the same time, ignore the comprehensive optimization of production efficiency, resulting in low monitoring and early warning accuracy.

[0006] To achieve the above objectives, the first aspect of the present invention provides a plant data management system and method based on digital twins, comprising: a data acquisition module, a data analysis module, and an equipment monitoring module;

[0007] Digital twin model module: used to build digital twin models of factory production equipment;

[0008] Data acquisition module: used to obtain simulation data of production equipment during simulation operation and environmental data at the corresponding time;

[0009] Data analysis module: calculates evaluation coefficients based on simulation data; wherein the evaluation data includes: failure evaluation coefficients and production evaluation coefficients; and,

[0010] Based on the environmental data, a correction model is constructed to correct the fault assessment coefficient to obtain the target fault coefficient; based on the pre-trained fault prediction model, the current simulation data is predicted to obtain the prediction result;

[0011] Equipment monitoring module: Determines whether to issue an early warning based on the prediction results.

[0012] Preferably, the construction of a digital twin model of the factory production equipment includes:

[0013] Obtain basic data of production equipment in the factory; basic data includes: equipment information and operating parameters;

[0014] Based on the basic data of the factory's production equipment, a digital twin model of the equipment is constructed using digital twin technology.

[0015] It should be noted that the device information is the basic static data for building a digital twin model, including the inherent attributes of the device, such as name, model, manufacturer, structural dimensions, material, design parameters (rated power, load capacity, etc.), technical documentation (CAD drawings, manuals), and installation location. This data usually does not change with the operating status and is mainly used for high-precision 3D modeling, equipment asset management, and maintenance traceability to ensure that the digital twin model is geometrically and logically consistent with the physical device.

[0016] Operating parameters are dynamic data generated by equipment in real time, reflecting its operating status. These include physical parameters (temperature, vibration, pressure), electrical data (current, voltage, energy consumption), performance indicators (efficiency, output), and fault alarm records. This data is collected in real time by sensors or industrial control systems (such as PLCs) and is used to monitor equipment health, predict failures, optimize operating strategies, and drive dynamic simulation and interaction of digital twin models, achieving intelligent operation and maintenance that synchronizes virtual and real systems.

[0017] Preferably, the obtaining of simulation data of the production equipment during simulation operation and environmental data at the corresponding time includes:

[0018] Obtain production tasks for pre-production products and run simulations based on the production tasks using the digital twin model of the equipment;

[0019] Collect simulation data of the equipment in the simulation process in real time; the simulation data includes: operation data and production data; operation data includes: equipment temperature and equipment vibration frequency; production data includes: production efficiency and equipment energy consumption;

[0020] The environmental data at the corresponding moment when the production equipment is running in simulation is acquired through data sensors; wherein the environmental data includes: ambient temperature and ambient humidity.

[0021] It should be noted that the production task of the pre-production product refers to a virtual production plan formulated based on product design requirements (such as specifications, quantity, and process parameters) before actual production. It simulates the equipment operation process, resource allocation, and production timing through a digital twin model to verify equipment performance, process feasibility, and production rhythm, ensuring that potential problems (such as equipment overload, excessive energy consumption, or efficiency bottlenecks) are identified before formal production, thereby optimizing parameters and reducing trial and error costs.

[0022] Preferably, the calculation of the evaluation coefficient based on the simulation data includes:

[0023] Calculate the fault assessment coefficient based on the operating data;

[0024] The production assessment coefficient is calculated based on the production data.

[0025] Preferably, the calculation of the fault assessment coefficient based on the operating data includes:

[0026] The temperature of the device is marked as SW, and the vibration frequency of the device is marked as SZ;

[0027] The fault assessment coefficient is calculated using the formula: P1 = α1 × e^[(SW-ZSW)^2 / (ZSW^2)] + α2 × ln[tanh(SZ)+1]; where P1 is the fault assessment coefficient, ZSW is the optimal operating temperature of the equipment, tanh(*) is the hyperbolic tangent function, and α1 and α2 are weight coefficients.

[0028] It should be noted that the weight coefficients α1 and α2 are set by those skilled in the art based on experience.

[0029] Preferably, the production evaluation coefficient is calculated based on the production data, including:

[0030] Mark production efficiency as V and equipment energy consumption as N;

[0031] The production evaluation coefficient is calculated using the formula: P2 = β1×e^[|V-ZV| / ln(ZV+1)]+β2×e^(N / (N+1)); where P2 is the production evaluation coefficient, ZV is the production efficiency required in the production task, and β1 and β2 are weight coefficients.

[0032] It should be noted that the weight coefficients β1 and β2 are set by those skilled in the art based on experience.

[0033] Preferably, the correction model is constructed based on the environmental data to correct the fault assessment coefficient, including:

[0034] The environmental data is used as the independent variable, and the ambient temperature in the environmental data is marked as W and the ambient humidity is marked as H;

[0035] The modified fault assessment coefficient is used as the dependent variable and is labeled as the target fault coefficient MP;

[0036] Fit the independent variables and dependent variables by polynomial fitting to build a revised model;

[0037] The correction model is specifically:

[0038] MP = P1 × (1 + θ1 × ln (e^(|W-ZW| / ZW)) + θ2 × ln (e^(|H-ZH| / ZH))); where ZW is the optimum ambient temperature for the device, ZH is the optimum ambient humidity for the device, and θ1 and θ2 are weight coefficients.

[0039] The target fault coefficient is obtained by inputting the currently calculated fault assessment coefficient P1 and environmental data into the correction model.

[0040] It should be noted that the weight coefficients θ1 and θ2 are set by those skilled in the art based on experience.

[0041] Preferably, the predicting of the current simulation data based on the pre-trained fault prediction model includes:

[0042] Based on the fault prediction model, the simulation data collected in real time is predicted to obtain the prediction results; wherein the prediction results refer to the target fault coefficient and production evaluation coefficient at the prediction time;

[0043] The training method of the fault prediction model is:

[0044] A1: Extract environmental data, simulation data, and historical data of corresponding target failure coefficients and production assessment coefficients;

[0045] A2: Integrate the environmental data, simulation data, and corresponding target failure coefficients and production evaluation coefficients at the initial moment in the historical data into standard input data; integrate the target failure coefficients and production evaluation coefficients at the prediction moment into standard output data;

[0046] A3: Train the artificial intelligence model based on standard input data and standard output data to obtain a fault prediction model.

[0047] It should be noted that the initial moment refers to the current moment that needs to be predicted, and the prediction moment refers to a certain moment in the future that needs to be predicted, which is specifically determined by the actual prediction needs.

[0048] Preferably, the determining whether to issue an early warning based on the prediction result includes:

[0049] Determine whether the target failure coefficient at the prediction moment is greater than the preset failure coefficient threshold; if yes, generate equipment failure warning information and send it to the client; if not, continue monitoring and judgment;

[0050] Determine whether the production evaluation coefficient at the prediction moment is greater than the preset production coefficient threshold; if so, generate production warning information and send it to the client; if not, continue monitoring and judgment.

[0051] It should be noted that the preset failure coefficient threshold and the preset production coefficient threshold are both set by those skilled in the art based on experience.

[0052] A second aspect of the present invention provides a plant data management method based on digital twins, comprising:

[0053] Step 1: Build a digital twin model of the factory's production equipment;

[0054] Step 2: Obtain simulation data of the production equipment during simulation operation and environmental data at the corresponding time;

[0055] Step 3: Calculate the evaluation coefficient based on the simulation data;

[0056] Step 4: Build a correction model based on environmental data to correct the fault assessment coefficient and obtain the target fault coefficient;

[0057] Step 5: Predict the current simulation data based on the pre-trained fault prediction model to obtain the prediction results;

[0058] Step 6: Determine whether to issue an early warning based on the prediction results.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The present invention constructs a digital twin model to comprehensively integrate the basic data and operating parameters of the factory's production equipment, thereby realizing accurate assessment of the equipment status, fully considering the coupling effect of environmental variables on faults, and effectively reducing the risk of misjudgment due to environmental fluctuations; the present invention not only focuses on the assessment of equipment failures, but also comprehensively optimizes production efficiency by introducing production assessment coefficients, making monitoring and early warning more comprehensive and accurate; by utilizing the data acquisition module to obtain simulation data and environmental data, and using the data analysis module for in-depth analysis, a correction model is constructed based on environmental data to correct the fault assessment coefficient, thereby further improving the accuracy of the assessment; at the same time, the pre-trained fault prediction model can predict the current simulation data, predict equipment failures and production anomalies in advance, and provide a valuable time window for equipment maintenance and production scheduling; through the equipment monitoring module, timely judgment and generation of early warning information based on the prediction results are sent to the client, realizing intelligent equipment monitoring and management, greatly improving the efficiency and reliability of factory operations, and reducing the workload and cost of manual inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0062] Figure 1 Schematic diagram of system modules according to an embodiment of the present invention;

[0063] Figure 2 Schematic diagram of the method steps of an embodiment of the present invention. DETAILED DESCRIPTION

[0064] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0065] See also Figure 1 , an embodiment of a first aspect of the present invention provides a plant data management system based on digital twins, comprising: a data acquisition module, a data analysis module and an equipment monitoring module;

[0066] Digital twin model module: used to build digital twin models of factory production equipment;

[0067] Data acquisition module: used to obtain simulation data of production equipment during simulation operation and environmental data at the corresponding time;

[0068] Data analysis module: calculates evaluation coefficients based on simulation data; wherein the evaluation data includes: failure evaluation coefficients and production evaluation coefficients; and,

[0069] Based on the environmental data, a correction model is constructed to correct the fault assessment coefficient to obtain the target fault coefficient; based on the pre-trained fault prediction model, the current simulation data is predicted to obtain the prediction result;

[0070] Equipment monitoring module: Determines whether to issue an early warning based on the prediction results.

[0071] Build a digital twin model of factory production equipment, including:

[0072] Obtain basic data of production equipment in the factory; basic data includes: equipment information and operating parameters;

[0073] Based on the basic data of the factory's production equipment, a digital twin model of the equipment is constructed using digital twin technology.

[0074] Get the simulation data of the production equipment during simulation and the corresponding environmental data at the time, including:

[0075] Obtain production tasks for pre-production products and run simulations based on the production tasks using the digital twin model of the equipment;

[0076] Collect simulation data of the equipment in the simulation process in real time; the simulation data includes: operation data and production data; operation data includes: equipment temperature and equipment vibration frequency; production data includes: production efficiency and equipment energy consumption;

[0077] The environmental data at the corresponding moment when the production equipment is running in simulation is acquired through data sensors; wherein the environmental data includes: ambient temperature and ambient humidity.

[0078] The evaluation coefficients are calculated based on simulation data, including:

[0079] Calculate the fault assessment coefficient based on the operating data;

[0080] The production assessment coefficient is calculated based on the production data.

[0081] The fault assessment coefficient is calculated based on the operating data, including:

[0082] The temperature of the device is marked as SW, and the vibration frequency of the device is marked as SZ;

[0083] The fault assessment coefficient is calculated using the formula: P1 = α1 × e^[(SW-ZSW)^2 / (ZSW^2)] + α2 × ln[tanh(SZ)+1]; where P1 is the fault assessment coefficient, ZSW is the optimal operating temperature of the equipment, tanh(*) is the hyperbolic tangent function, and α1 and α2 are weight coefficients.

[0084] The production assessment coefficient is calculated based on production data, including:

[0085] Mark production efficiency as V and equipment energy consumption as N;

[0086] The production evaluation coefficient is calculated using the formula: P2 = β1×e^[|V-ZV| / ln(ZV+1)]+β2×e^(N / (N+1)); where P2 is the production evaluation coefficient, ZV is the production efficiency required in the production task, and β1 and β2 are weight coefficients.

[0087] A correction model is constructed based on environmental data to correct the fault assessment coefficient, including:

[0088] The environmental data is used as the independent variable, and the ambient temperature in the environmental data is marked as W and the ambient humidity is marked as H;

[0089] The modified fault assessment coefficient is used as the dependent variable and is labeled as the target fault coefficient MP;

[0090] Fit the independent variables and dependent variables by polynomial fitting to build a revised model;

[0091] The correction model is specifically:

[0092] MP = P1 × (1 + θ1 × ln (e^(|W-ZW| / ZW)) + θ2 × ln (e^(|H-ZH| / ZH))); where ZW is the optimum ambient temperature for the device, ZH is the optimum ambient humidity for the device, and θ1 and θ2 are weight coefficients.

[0093] The target fault coefficient is obtained by inputting the currently calculated fault assessment coefficient P1 and environmental data into the correction model.

[0094] Predict the current simulation data based on the pre-trained fault prediction model, including:

[0095] Based on the fault prediction model, the simulation data collected in real time is predicted to obtain the prediction results; wherein the prediction results refer to the target fault coefficient and production evaluation coefficient at the prediction time;

[0096] The training method of the fault prediction model is:

[0097] A1: Extract environmental data, simulation data, and historical data of corresponding target failure coefficients and production assessment coefficients;

[0098] A2: Integrate the environmental data, simulation data, and corresponding target failure coefficients and production evaluation coefficients at the initial moment in the historical data into standard input data; integrate the target failure coefficients and production evaluation coefficients at the prediction moment into standard output data;

[0099] A3: Train the artificial intelligence model based on standard input data and standard output data to obtain a fault prediction model.

[0100] Determine whether to issue an early warning based on the prediction results, including:

[0101] Determine whether the target failure coefficient at the prediction moment is greater than the preset failure coefficient threshold; if yes, generate equipment failure warning information and send it to the client; if not, continue monitoring and judgment;

[0102] Determine whether the production evaluation coefficient at the prediction moment is greater than the preset production coefficient threshold; if so, generate production warning information and send it to the client; if not, continue monitoring and judgment.

[0103] For example, an auto parts factory uses an injection molding machine to produce plastic housings. The equipment parameters are as follows:

[0104] Optimum working temperature (ZSW): 180℃;

[0105] Optimum ambient temperature (ZW): 25°C, optimal humidity (ZH): 50%;

[0106] Pre-production task requirements: production efficiency (ZV) = 100 pieces / hour, unit energy consumption is less than or equal to 0.8 kWh / piece.

[0107] 1. Digital twin model construction and simulation;

[0108] Step 1: Basic data collection;

[0109] Equipment information: Injection molding machine model JZ-200, rated power 50kW, design pressure 120MPa, CAD drawings imported to generate 3D models.

[0110] Operating parameters: Real-time temperature, vibration frequency, and energy consumption data are collected through PLC and sensors (temperature sensor, vibration sensor, and electricity meter).

[0111] Step 2: Pre-production task simulation;

[0112] Virtual production plan: simulate the production of 5,000 shells, with the process parameters set to an injection pressure of 110 MPa and a cooling time of 15 seconds.

[0113] Simulation output data:

[0114] Operating data: temperature SW = 185 ° C, vibration frequency SZ = 45 Hz;

[0115] Production data: production efficiency V = 95 pieces / hour, energy consumption N = 0.85 kWh / piece;

[0116] Environmental data: workshop temperature W = 28 ° C, humidity H = 55% (collected in real time through temperature and humidity sensors).

[0117] 2. Calculation of dual-dimensional evaluation coefficients;

[0118] (1) Calculation of fault assessment coefficient (P1):

[0119] The fault assessment coefficient is calculated using the formula P1 = α1 × e^[(SW-ZSW)^2 / (ZSW^2)] + α2 × ln[tanh(SZ)+1];

[0120] Assume that the weight coefficients α1 = 0.6, α2 = 0.4;

[0121] Substituting the data, we get the fault assessment coefficient P1≈0.878;

[0122] (2) Calculation of production evaluation coefficient (P2);

[0123] The production assessment coefficient is calculated using the formula P2 = β1 × e^[|V-ZV| / ln(ZV+1)] + β2 × e^(N / (N+1));

[0124] Assume that the weight coefficients β1 = 0.7, β2 = 0.3;

[0125] Substituting the data, we obtain the production evaluation coefficient P2≈2.554.

[0126] 3. Environmental correction and target failure coefficient (MP);

[0127] Corrected model formula: MP = P1 × (1 + θ1 × ln (e^(|W-ZW| / ZW)) + θ2 × ln (e^(|H-ZH| / ZH)));

[0128] Assume that the weight coefficients θ1 = 0.6, θ2 = 0.4;

[0129] Substituting the data, we get the target failure coefficient MP≈0.976;

[0130] 4. Fault prediction and early warning decision-making;

[0131] (1) AI prediction model input and output:

[0132] Input data (current moment):

[0133] Environmental data: W = 28 ° C, H = 55%;

[0134] Simulation data: SW = 185°C, SZ = 45 Hz, V = 95, N = 0.85;

[0135] Evaluation coefficient: MP = 0.976, P2 = 2.554;

[0136] Output prediction (next 2 hours):

[0137] Prediction MP = 1.072;

[0138] Prediction P2 = 2.8;

[0139] (2) Warning trigger:

[0140] Assume that the preset failure coefficient threshold is 1 and the preset production coefficient threshold is 2.6;

[0141] Fault warning: Since the MP prediction value (1.072) is greater than the preset fault coefficient threshold (1), an early warning is sent to the maintenance department, prompting "Risk of overheating of injection molding machine bearings, it is recommended to check the cooling system."

[0142] Production warning: Since the P2 predicted value (2.8) is greater than the preset production coefficient threshold (2.6), an early warning is sent to the production scheduling center, prompting "Energy consumption exceeds the standard, it is recommended to adjust process parameters or reduce the production pace."

[0143] 5. Actual optimization effect;

[0144] Maintenance response: Inspection revealed insufficient cooling water flow. After cleaning the pipeline, the temperature returned to 180°C and the failure coefficient MP dropped to 0.82.

[0145] Production adjustments: The injection pressure was reduced from 110 MPa to 105 MPa, the unit energy consumption was reduced to 0.78 kWh / piece, and the production efficiency was increased to 98 pieces / hour.

[0146] Comprehensive benefits: Avoid downtime losses caused by bearing overheating and reduce energy costs.

[0147] This example uses a digital twin model to dynamically simulate the injection molding machine production process, incorporates environmental data to correct fault assessment coefficients, and leverages an AI-powered predictive model to provide a two-hour advance warning of equipment overheating and excessive energy consumption. The system pinpoints issues using quantitative metrics (MP and P2), guides maintenance and production optimization, and ultimately reduces downtime losses and energy costs, demonstrating the significant advantages of this approach in improving the accuracy and efficiency of factory operations and maintenance.

[0148] See Figure 2The second embodiment of the present invention provides a plant data management method based on digital twins, including:

[0149] Step 1: Build a digital twin model of the factory's production equipment;

[0150] Step 2: Obtain simulation data of the production equipment during simulation operation and environmental data at the corresponding time;

[0151] Step 3: Calculate the evaluation coefficient based on the simulation data;

[0152] Step 4: Build a correction model based on environmental data to correct the fault assessment coefficient and obtain the target fault coefficient;

[0153] Step 5: Predict the current simulation data based on the pre-trained fault prediction model to obtain the prediction results;

[0154] Step 6: Determine whether to issue an early warning based on the prediction results.

[0155] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0156] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A plant data management system based on digital twins, characterized by: include: Data acquisition module, data analysis module and equipment monitoring module; Digital twin model module: used to build digital twin models of factory production equipment; Data acquisition module: used to obtain simulation data of production equipment during simulation operation and environmental data at the corresponding time; Data analysis module: calculates evaluation coefficients based on simulation data; wherein the evaluation data includes: failure evaluation coefficients and production evaluation coefficients; and, Based on the environmental data, a correction model is constructed to correct the fault assessment coefficient to obtain the target fault coefficient; based on the pre-trained fault prediction model, the current simulation data is predicted to obtain the prediction result; Equipment monitoring module: Determine whether to issue an early warning based on the prediction results.

2. A plant data management system based on digital twins according to claim 1, characterized in that: The construction of a digital twin model of factory production equipment includes: Obtain basic data of production equipment in the factory; basic data includes: equipment information and operating parameters; Based on the basic data of the factory's production equipment, a digital twin model of the equipment is constructed using digital twin technology.

3. The plant data management system based on digital twin according to claim 1, characterized in that: The obtaining of simulation data of the production equipment during simulation operation and environmental data at the corresponding time includes: Obtain production tasks for pre-production products and run simulations based on the production tasks using the digital twin model of the equipment; Collect simulation data of the equipment in the simulation process in real time; the simulation data includes: operation data and production data; operation data includes: equipment temperature and equipment vibration frequency; production data includes: production efficiency and equipment energy consumption; The environmental data at the corresponding moment when the production equipment is running in simulation is acquired through data sensors; wherein the environmental data includes: ambient temperature and ambient humidity.

4. The plant data management system based on digital twin according to claim 1, characterized in that: The evaluation coefficient is calculated based on the simulation data, including: Calculate the fault assessment coefficient based on the operating data; The production assessment coefficient is calculated based on the production data.

5. The plant data management system based on digital twin according to claim 4 is characterized in that: The fault assessment coefficient is calculated based on the operating data, including: The temperature of the device is marked as SW, and the vibration frequency of the device is marked as SZ; The fault assessment coefficient is calculated using the formula: P1 = α1 × e^[(SW-ZSW)^2 / (ZSW^2)] + α2 × ln[tanh(SZ)+1]; where P1 is the fault assessment coefficient, ZSW is the optimal operating temperature of the equipment, tanh(*) is the hyperbolic tangent function, and α1 and α2 are weight coefficients.

6. The plant data management system based on digital twin according to claim 4 is characterized in that: The production assessment coefficient is calculated based on the production data, including: Mark production efficiency as V and equipment energy consumption as N; The production evaluation coefficient is calculated using the formula: P2 = β1×e^[|V-ZV| / ln(ZV+1)]+β2×e^(N / (N+1)); where P2 is the production evaluation coefficient, ZV is the production efficiency required in the production task, and β1 and β2 are weight coefficients.

7. The plant data management system based on digital twin according to claim 1, characterized in that: The correction model is constructed based on the environmental data to correct the fault assessment coefficient, including: The environmental data is used as the independent variable, and the ambient temperature in the environmental data is marked as W and the ambient humidity is marked as H; The modified fault assessment coefficient is used as the dependent variable and is labeled as the target fault coefficient MP; Fit the independent variables and dependent variables by polynomial fitting to build a revised model; The correction model is specifically: MP = P1 × (1 + θ1 × ln (e^(|W-ZW| / ZW)) + θ2 × ln (e^(|H-ZH| / ZH))); where ZW is the optimum ambient temperature for the device, ZH is the optimum ambient humidity for the device, and θ1 and θ2 are weight coefficients. The target fault coefficient is obtained by inputting the currently calculated fault assessment coefficient P1 and environmental data into the correction model.

8. The plant data management system based on digital twin according to claim 1, characterized in that: The fault prediction model obtained by pre-training is used to predict the current simulation data, including: Based on the fault prediction model, the simulation data collected in real time is predicted to obtain the prediction results; wherein the prediction results refer to the target fault coefficient and production assessment coefficient at the prediction time; The training method of the fault prediction model is: A1: Extract environmental data, simulation data, and historical data of corresponding target failure coefficients and production assessment coefficients; A2: Integrate the environmental data, simulation data, and corresponding target failure coefficients and production evaluation coefficients at the initial moment in the historical data into standard input data; integrate the target failure coefficients and production evaluation coefficients at the prediction moment into standard output data; A3: Train the artificial intelligence model based on standard input data and standard output data to obtain a fault prediction model.

9. The plant data management system based on digital twin according to claim 1, characterized in that: The determining whether to issue an early warning based on the prediction result includes: Determine whether the target failure coefficient at the prediction moment is greater than the preset failure coefficient threshold; if yes, generate equipment failure warning information and send it to the client; if not, continue monitoring and judgment; Determine whether the production evaluation coefficient at the prediction moment is greater than the preset production coefficient threshold; if so, generate production warning information and send it to the client; if not, continue monitoring and judgment.

10. A plant data management method based on digital twins, applied to a plant data management system based on digital twins according to any one of claims 1 to 9, characterized in that: include: Step 1: Build a digital twin model of the factory's production equipment; Step 2: Obtain simulation data of the production equipment during simulation operation and environmental data at the corresponding time; Step 3: Calculate the evaluation coefficient based on the simulation data; Step 4: Build a correction model based on environmental data to correct the fault assessment coefficient and obtain the target fault coefficient; Step 5: Predict the current simulation data based on the pre-trained fault prediction model to obtain the prediction results; Step 6: Determine whether to issue an early warning based on the prediction results.

Images