Intelligent daily chemical production line online control system

Abstract

The invention relates to the technical field of program control systems, in particular to an intelligent daily chemical production line online control system. The system comprises an equipment digital twin modeling module, a process parameter optimization module, a modular production unit adjustment module, an automatic scheduling and Internet of Things transmission module, an equipment monitoring and data analysis module and a fault prediction and remote operation and maintenance module. Acquiring equipment real-time operation state data of the intelligent daily chemical production line, and performing digital twinborn modeling based on the equipment real-time operation state data to obtain an equipment digital twinborn model; and performing simulation optimization calculation on each process parameter based on the digital twin model of the equipment to obtain stable-state process parameter set data of the equipment. Through technological parameter optimization, modular production unit adjustment, equipment monitoring and fault prediction, efficient dynamic optimization of the production process is realized, the flexibility, stability and fault prevention capability of the production line are improved, the production cost is reduced, and the production efficiency is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of program control systems, and particularly to an online control system for an intelligent daily chemical production line. Background Art

[0002] The existing intelligent daily chemical production line has insufficient process parameter optimization ability, resulting in difficulty in improving production efficiency. Currently, the process parameters of daily chemical production mainly rely on empirical settings and lack an efficient simulation optimization mechanism. Since the optimization calculation of process parameters mainly relies on manual experience or simple rule settings and fails to fully combine the real-time status data of equipment for simulation optimization, parameter adjustment lags behind and it is difficult to achieve dynamic optimization. In terms of automated scheduling and Internet of Things data transmission, traditional daily chemical production lines rely on fixed scheduling rules and are difficult to meet the needs of flexible production. The current production scheduling method mainly relies on a preset production plan and lacks an intelligent task allocation and dynamic optimization mechanism. Facing the rapid changes in market demand, production lines are often unable to adjust production plans in a timely manner, resulting in a decline in production efficiency. The ability of fault prediction and remote operation and maintenance is weak, leading to high equipment maintenance costs and difficulty in ensuring production continuity. In the absence of intelligent predictive maintenance means, equipment failures are often discovered when the problems are serious, resulting in production interruptions and increased maintenance costs. Summary of the Invention

[0003] Based on this, it is necessary for the present invention to provide an online control system for an intelligent daily chemical production line to solve at least one of the above technical problems.

[0004] To achieve the above object, an online control system for an intelligent daily chemical production line includes the following modules:

[0005] An equipment digital twin modeling module, configured to obtain the real-time operation status data of the equipment of the intelligent daily chemical production line and perform digital twin modeling based on the real-time operation status data of the equipment to obtain an equipment digital twin model;

[0006] A process parameter optimization module, configured to perform spatio-temporal graph neural network simulation optimization calculation on each process parameter based on the equipment digital twin model and perform multi-objective optimization simulation calculation to obtain data of a set of stable-state process parameters of the equipment;

[0007] A modular production unit adjustment module, configured to optimize the scheduling strategy of the modular production unit based on the data of the set of stable-state process parameters of the equipment and perform dynamic characteristic modeling of the modular production unit to obtain optimized data of the modular production unit;

[0008] An automated scheduling and Internet of Things transmission module, configured to perform automated scheduling on the optimized data of the modular production unit and perform Internet of Things remote data transmission on the automated scheduling result to obtain equipment remote production scheduling instruction data;

[0009] The equipment monitoring and data analysis module is used to analyze the temporal and spatial evolution trend of the intelligent daily chemical production line based on the equipment remote production scheduling instruction data, and to perform real-time equipment monitoring and generate equipment operation analysis data;

[0010] The fault prediction and remote operation and maintenance module is used to perform spatiotemporal pattern recognition of equipment health status based on equipment operation analysis data, use the spatiotemporal pattern recognition results to predict equipment faults and design predictive maintenance plans, and remotely operate and maintain the intelligent daily chemical production line based on the predictive maintenance plan design results to obtain an optimized production control system.

[0011] The online control system of the intelligent daily chemical production line proposed in the present invention integrates equipment digital twin modeling, process parameter optimization, modular production unit adjustment, automated scheduling and Internet of Things data transmission, equipment monitoring and data analysis, as well as fault prediction and remote operation and maintenance modules, which significantly improves the overall dynamic response capability of the production line and the accuracy of process parameter optimization. It uses real-time collected equipment operation status data to build a high-fidelity digital twin to achieve accurate reproduction of equipment status information, and determines process parameters based on detailed spatiotemporal simulation calculations. It generates a steady-state process parameter set through spatiotemporal graph neural network and multi-objective optimization algorithm, thereby effectively avoiding the lag and inefficiency of traditional empirical settings. At the same time, modular production unit adjustment is implemented. The scheduling strategy optimization and dynamic characteristic modeling based on real-time data can meet the needs of flexible production. The automated scheduling and Internet of Things data transmission module ensures that the scheduling instructions are issued immediately through fixed communication protocols and real-time data transmission mechanisms. The equipment monitoring and data analysis module ensures real-time feedback of operating status information through rigorous spatiotemporal evolution trend analysis and continuous monitoring. The fault prediction and remote operation and maintenance module relies on strict spatiotemporal pattern recognition and predictive maintenance solution design to achieve early warning of equipment failures and remote operation and maintenance operations, thereby reducing equipment failure rate and maintenance costs, ensuring production continuity, and realizing overall refined management and dynamic optimization of online control of the production line, significantly improving production efficiency and system operation stability.

[0012] Preferably, the device digital twin modeling module includes the following functions:

[0013] Obtain the real-time operation status data of the equipment of the intelligent daily chemical production line, and perform high-precision modeling of the geometric and physical characteristics of the equipment based on the real-time operation status data of the equipment to obtain the model data of the physical characteristics of the equipment;

[0014] Perform equipment dynamic behavior simulation based on equipment physical characteristic model data to obtain equipment dynamic behavior data, wherein the equipment dynamic behavior simulation includes vibration behavior simulation, thermodynamic behavior simulation, electromagnetic behavior simulation and fluid dynamic behavior simulation;

[0015] Optimize the virtual mapping of device operation parameters using the implicit representation method of neural radiance fields with device dynamic behavior data to obtain device virtual mapping data;

[0016] Perform electromagnetic field effect, material fatigue analysis, and fluid interaction analysis on the device virtual mapping data, and construct a multi-physical field coupled digital twin environment based on the results of electromagnetic field effect, material fatigue analysis, and fluid interaction analysis to obtain device digital twin environment data;

[0017] Perform fusion analysis on the device real-time operation state data based on the device digital twin environment data to obtain a preliminary device digital twin model;

[0018] Perform anomaly detection on the preliminary device digital twin model, and optimize the model parameters based on the anomaly detection results to obtain a device digital twin model.

[0019] In this invention, by acquiring the real-time operation state data of intelligent daily chemical production line equipment and performing high-precision modeling of the geometric and physical characteristics of the equipment based on this, the accuracy of the equipment physical characteristic model can be ensured, and the ability of the model to reflect the real operation state of the equipment can be improved. Based on this physical characteristic model, perform device dynamic behavior simulation, so that the vibration, thermodynamics, electromagnetics, and fluid dynamics characteristics of the equipment are comprehensively simulated, thereby accurately predicting the dynamic performance of the equipment under different working conditions, and enhancing the safety and stability of equipment operation. Use device dynamic behavior data to optimize the implicit representation method of neural radiance fields, making the virtual mapping of device operation parameters more accurate, and providing an efficient parameter adjustment basis for device optimization control. By performing electromagnetic field effect, material fatigue analysis, and fluid interaction analysis on the virtual mapping data, construct a multi-physical field coupled digital twin environment, making the prediction of the device operation state more comprehensive and the optimization plan more targeted. On this basis, perform fusion analysis on the device real-time operation state data, establish an accurate device digital twin model, thereby improving the comprehensive perception ability of the device operation state. Further perform anomaly detection on the digital twin model and optimize the model parameters, which can enhance the self-adaptability and anomaly recognition ability of the model, improve the accuracy of device predictive maintenance, reduce the device failure rate, extend the service life of the device, and ultimately enhance the intelligent management level and production efficiency of the production line.

[0020] Preferably, the optimization of the virtual mapping of device operation parameters using the implicit representation method of neural radiance fields with device dynamic behavior data includes:

[0021] Extract multi-dimensional features of the device operation state based on the device dynamic behavior data to obtain device dynamic behavior feature data;

[0022] Perform implicit representation encoding on the device dynamic behavior feature data, and construct a preliminary neural radiance field model based on the implicit representation encoding result to obtain initial device implicit representation model data;

[0023] Based on the initial device implicit representation model data, perform adaptive weight allocation and regularization strategy adjustment to obtain the structure-optimized device implicit representation model data;

[0024] Perform backpropagation training on the structure-optimized implicit representation model data, and based on the backpropagation training results, perform cross-validation and model parameter adjustment to obtain the fine-optimized device implicit representation model data;

[0025] Based on the fine-optimized device implicit representation model data, perform virtual mapping of device operation parameters, and perform error feedback correction on the results of the virtual mapping of device operation parameters to obtain device virtual mapping data.

[0026] The present invention can fully exploit the core characteristics of the device operation state by performing multi-dimensional feature extraction on the device dynamic behavior data, improving the recognition and modeling accuracy of the complex dynamic behavior of the device. Using the implicit representation coding method, the high-dimensional dynamic behavior feature data can be compressed into a low-dimensional latent space, reducing data redundancy while enhancing the information expression ability, thereby providing more compact and efficient input data for the construction of the neural radiance field model. Based on the initial implicit representation model data, performing adaptive weight allocation and regularization strategy adjustment helps to improve the model's adaptability to different features, suppress the overfitting problem, and enhance the generalization ability and stability of the model. By optimizing the implicit representation model through backpropagation training and combining cross-validation and parameter adjustment, the convergence and accuracy of the model can be further enhanced, ensuring a more accurate mapping of the device operation state. Finally, during the device virtual mapping process, by combining the error feedback correction mechanism, the mapping result is closer to the real physical characteristics, improving the prediction accuracy and adaptive optimization ability of the model, providing more reliable data support for the intelligent optimization control of the device, thereby improving the device operation efficiency, reducing energy consumption, and enhancing the intelligent level of the production line.

[0027] Preferably, the process parameter optimization module includes the following functions:

[0028] Extract spatio-temporal dimension features from the device digital twin model to obtain process parameter spatio-temporal feature vector data;

[0029] Based on the process parameter spatio-temporal feature vector data, construct a spatio-temporal graph structure to obtain dynamic spatio-temporal relationship graph data;

[0030] Perform spatio-temporal graph convolutional network modeling on the dynamic spatio-temporal relationship graph data to obtain a spatio-temporal feature fusion model;

[0031] Use the spatio-temporal feature fusion model to perform multi-scale spatio-temporal attention calculation to obtain dynamic weight allocation matrix data;

[0032] Perform process parameter spatio-temporal evolution simulation based on the dynamic weight assignment matrix data to obtain parameter spatio-temporal response surface data;

[0033] Perform multi-objective genetic algorithm optimization based on the parameter spatio-temporal response surface data, and perform dynamic game theory conflict resolution on the optimization results of the multi-objective genetic algorithm to obtain the data of the stable-state process parameter set of the equipment.

[0034] Through spatio-temporal dimension feature extraction of the equipment digital twin model, the present invention can accurately capture the variation law of process parameters at different time and space scales, and improve the comprehensive perception ability of the process state. Constructing a spatio-temporal graph structure can effectively reveal the dynamic correlation between process parameters and enhance the analysis ability of complex process evolution, thereby providing high-quality data support for subsequent modeling. Using a spatio-temporal graph convolutional network for modeling enables the model to efficiently extract spatio-temporal relationship features and improve the understanding ability of the dynamic change trend of process parameters. Combining multi-scale spatio-temporal attention calculation can achieve accurate identification of key influencing factors and dynamic weight assignment, enhance the self-adaptability of the model, and improve the intelligent perception ability of the process optimization process. Performing process parameter spatio-temporal evolution simulation based on the dynamic weight assignment matrix enables the variation trend of process parameters under different working conditions to be visually expressed, thereby improving the prediction accuracy of complex process. Finally, through multi-objective genetic algorithm optimization combined with dynamic game theory conflict resolution, the competitive relationship between process parameters can be effectively balanced, the stable operation state of the equipment can be optimized, and the process parameters can reach the optimal configuration under various constraints, thereby improving production efficiency, reducing energy consumption, and enhancing the stability and intelligent level of the production line.

[0035] Preferably, the modular production unit adjustment module includes the following functions:

[0036] Perform matching of the operation parameters of the modular production unit based on the data of the stable-state process parameter set of the equipment to obtain the data of the production unit parameters of the equipment;

[0037] Perform optimization analysis of the collaborative operation characteristics of each modular production unit based on the initial production unit parameter data of the equipment to obtain the collaborative optimization data of the production units of the equipment;

[0038] Use the collaborative optimization data of the production units of the equipment to perform calculation of the adjustment of the process flow parameters of each modular production unit to obtain the process parameter data of the optimized production unit of the equipment;

[0039] Perform dynamic adaptation adjustment of the production capacity of the modular production unit based on the process parameter data of the optimized production unit of the equipment to obtain the data of the optimized modular production unit.

[0040] The present invention can ensure that each production unit accurately adjusts its operating parameters based on the optimized process parameters by matching the operating parameters of modular production units based on the data of the stable-state process parameter set of the equipment, thereby improving production efficiency, reducing energy consumption, and ensuring product consistency and quality. By optimizing the collaborative operation characteristics analysis of each modular production unit based on the initial production unit parameter data of the equipment, the collaborative effects between production units can be identified, providing an accurate optimization plan for the collaborative work of multiple modules and avoiding the situation where the improvement of a single production unit cannot bring about an overall benefit increase. By using the collaborative optimization data of the equipment production units to calculate the adjustment of the process flow parameters of each modular production unit, the process flows can be refined, ensuring that each module can operate efficiently and improving production capacity and production stability under the collaborative effects of the entire production line. By dynamically adapting and adjusting the production capacity of modular production units based on the optimized production unit process parameter data of the equipment, different demand changes and production loads during the production process can be flexibly responded to, ensuring that the production line can quickly adapt and operate efficiently when facing different production demands, thereby maximizing production benefits and the resource utilization rate of the system.

[0041] Preferably, the automated scheduling and IoT transmission module includes the following functions:

[0042] Decompose the production tasks for the optimized modular production unit data to obtain the equipment production task decomposition data;

[0043] Calculate the production resource allocation based on the equipment production task decomposition data to obtain the equipment production resource allocation data;

[0044] Optimize and adjust the production scheduling plan of the intelligent daily chemical production line using the equipment production resource allocation data to obtain the equipment production scheduling optimization data;

[0045] Monitor and verify the scheduling execution status based on the equipment production scheduling optimization data to obtain the equipment automated scheduling result data;

[0046] Conduct IoT data transmission based on the equipment automated scheduling result data to generate remote instructions to obtain the equipment remote production scheduling instruction data.

[0047] By decomposing production tasks based on the optimized modular production unit data, the present invention can refine complex production tasks into more operable units, making the production process clearer and more systematic, and enhancing the controllability and monitorability of the production process. Based on the production task decomposition data of the equipment, calculating the production resource allocation can maximize the resource utilization rate while ensuring the rational allocation of resources, thereby reducing resource waste and enhancing the overall efficiency of the production line. Using the production resource allocation data of the equipment to optimize and adjust the production scheduling plan for the intelligent daily chemical production line can flexibly adjust the production plan in real time according to the resource allocation situation, ensure the optimization of production scheduling and adapt to dynamic changes, and enhance the flexibility and response speed of the production line. Based on the optimized data of the equipment production scheduling, monitoring and verifying the scheduling execution status helps to track the execution of the production process in real time, discover and correct problems in a timely manner, and ensure the efficient execution of the scheduling plan. Based on the automated scheduling result data of the equipment, transmitting Internet of Things data to generate remote instructions can realize the remote control and monitoring of the production line, enhance the flexibility and management efficiency of production, provide real-time response capabilities for enterprises in dealing with emergencies during the production process, and thus enhance the intelligence and automation level of the production line.

[0048] Preferably, the equipment monitoring and data analysis module includes the following functions:

[0049] Collecting the operation status of each modular production unit of the intelligent daily chemical production line based on the equipment remote production scheduling instruction data to obtain the equipment production unit operation status data;

[0050] Detecting abnormalities in key operation parameters of the equipment production unit operation status data to obtain equipment abnormality detection data;

[0051] Analyzing the trend of the health status of key equipment components based on the equipment abnormality detection data to obtain equipment health status analysis data;

[0052] Analyzing the production efficiency of the equipment and evaluating its stability based on the equipment health status analysis data to obtain equipment operation analysis data.

[0053] The present invention collects the operation status of each modular production unit of the intelligent daily chemical production line based on the remote production scheduling instruction data of the equipment, can obtain the operation status data of each module of the production line in real time, ensure that every link in the production process can be accurately monitored, and help quickly identify potential problems. By detecting abnormal key operation parameters in the operation status data of the equipment production unit, abnormal fluctuations and potential faults in the equipment operation can be found in real time, preventing the problem from deteriorating further, thereby reducing the downtime and maintenance costs. Based on the equipment anomaly detection data, trend analysis of the health status of key equipment components is carried out, which can accurately evaluate the health status of key equipment components, predict the future health trend of the equipment, ensure that measures are taken in a timely manner for maintenance and repair, and avoid the risk of production interruption caused by sudden failures. Based on the equipment health status analysis data, equipment production efficiency analysis and stability evaluation are carried out, which can comprehensively evaluate the production efficiency and operation stability of the equipment, help managers optimize production scheduling, adjust operation strategies, improve the overall performance of the equipment, ensure that the production line remains efficient and stable during long-term operation, thereby maximizing production benefits and reducing the impact of production fluctuations.

[0054] Preferably, the equipment production efficiency analysis and stability evaluation based on the equipment health status analysis data include:

[0055] Perform time series analysis on the equipment health status analysis data to obtain equipment operation time characteristic data;

[0056] Calculate the production task completion rate based on the equipment operation time characteristic data to obtain equipment production task completion rate data;

[0057] Build an equipment spatio-temporal key feature model based on the equipment production task completion rate data to obtain equipment energy consumption analysis data;

[0058] Detect fluctuations in the operation stability of the equipment based on the equipment energy consumption analysis data to obtain equipment stability detection data;

[0059] Evaluate the load status of key equipment components based on the equipment stability detection data to obtain equipment key component load data;

[0060] Calculate the equipment production efficiency using the equipment production task completion rate data and the equipment key component load data to obtain equipment production efficiency analysis data;

[0061] Carry out spatio-temporal trend analysis of the long-term operation stability of the equipment based on the equipment production efficiency analysis data to obtain equipment operation analysis data.

[0062] Through time series analysis of the device health status analysis data, the present invention can accurately extract the long-term operation characteristics of the device, identify the time characteristics of different stages in the device operation process, so as to provide accurate data support for the subsequent calculation of the production task completion rate. In the process of calculating the production task completion rate, combining the time characteristic data can accurately measure the execution efficiency of the production unit, improve the matching degree of the production plan, and optimize the production scheduling strategy. Based on the production task completion rate data, the spatio-temporal key characteristics of the device are modeled, so that the energy consumption distribution of the device can be accurately characterized in the time and space dimensions, thereby revealing the energy consumption characteristics and energy efficiency bottlenecks of each production unit, and providing a basis for energy-saving optimization. On the basis of energy consumption analysis, the detection of the stability fluctuation of the device operation is carried out, which can accurately identify the stability change of the device under different working conditions, timely discover the abnormal operation fluctuation, and prevent the unplanned shutdown of the device caused by abnormal energy consumption or load fluctuation. By evaluating the load status of the key components based on the device stability detection data, the bearing capacity of the key components of the device can be evaluated in real time, and then the load distribution can be optimized, the service life of the key components of the device can be extended, and the overall reliability of the production line can be improved. By combining the production task completion rate data and the key component load data to calculate the production efficiency of the device, the resource utilization rate, energy efficiency ratio and device utilization rate in the production process can be comprehensively analyzed, providing a reliable basis for the optimization of production efficiency. Based on the long-term operation stability spatio-temporal trend analysis of the device production efficiency analysis data, the trend changes of the device in long-term operation can be revealed, the fluctuation trend of the production efficiency can be predicted in advance, and the production line can be ensured to maintain efficient and stable operation under different production loads and environmental change conditions.

[0063] Preferably, the fault prediction and remote operation and maintenance module includes the following functions:

[0064] Based on the device operation analysis data, a historical operation record trend model is established to obtain device operation trend data;

[0065] Based on the device operation trend data, a historical fault characteristic pattern matching analysis of the key components of the device is carried out to obtain device fault prediction data;

[0066] Based on the device fault prediction data, an optimization calculation of the maintenance strategy is carried out to obtain device predictive maintenance plan data;

[0067] Based on the device predictive maintenance plan data, remote operation and maintenance of the intelligent daily chemical production line is carried out to obtain an optimized production control system.

[0068] By performing historical operation record trend modeling based on device operation analysis data, the present invention can identify potential problems during the device operation, reveal the stability and performance change trends of the long-term device operation, and provide data support for subsequent fault prediction and maintenance decision-making. By performing historical fault feature pattern matching analysis on key components of the device based on device operation trend data, the fault patterns that occur in the device can be identified in advance, and the future fault risks can be predicted by comparing historical data, thereby avoiding production interruptions caused by sudden failures. By performing maintenance strategy optimization calculation based on device fault prediction data, a scientific and reasonable predictive maintenance plan can be formulated to ensure that the device is repaired in time before a fault occurs, reduce the occurrence of sudden failures, and improve the reliability of the device and the continuous operation ability of the production line. By remotely operating and maintaining the intelligent daily chemical production line based on the device predictive maintenance plan data, the health status of the device can be grasped in real time, the maintenance plan can be adjusted in time, the stability and operation efficiency of the production system are improved, and finally the optimization of the production control system is realized, enabling the device to operate in an efficient and stable state, reducing the maintenance cost and downtime, and ensuring the smooth progress of the production task.

[0069] Preferably, the historical fault feature pattern matching analysis on key components of the device based on device operation trend data includes:

[0070] Extract historical fault records from the device operation trend data and classify the fault patterns to obtain device fault pattern classification data;

[0071] Based on the device fault pattern classification data, perform statistical analysis on the characteristic parameters of each fault pattern to obtain device fault characteristic parameter data;

[0072] Based on the device fault characteristic parameter data, perform device key component operation state characteristic matching calculation to obtain device fault characteristic matching data;

[0073] Use the device fault characteristic matching data to calculate the probability of occurrence of faults in the key components of the device to obtain device fault probability prediction data;

[0074] Based on the device fault probability prediction data, perform device operation state abnormal trend analysis to obtain device abnormal trend analysis data;

[0075] Use the device abnormal trend analysis data to perform fault risk level assessment to obtain device fault risk level data;

[0076] Integrate the device fault probability prediction data and the device fault risk level data to obtain device fault prediction data.

[0077] By extracting historical fault records from the equipment operation trend data and classifying the fault modes, the present invention can deeply analyze the historical fault conditions of the equipment, identify the characteristics of different fault modes, and provide basic data for subsequent fault prediction. Based on the classified data of the equipment fault modes, statistical analysis of the characteristic parameters of each fault mode helps to reveal the specific manifestations and key characteristics of different fault modes, assists in accurately evaluating the potential risks of the equipment operation, and provides accurate parameter basis for subsequent fault diagnosis and prevention. Based on the data of the equipment fault characteristic parameters, calculating the matching of the operating state characteristics of the key components of the equipment can effectively identify the abnormal states of the key components in actual operation, discover fault signs in advance, and thus avoid equipment damage and production stagnation. Using the equipment fault matching data to calculate the occurrence probability of the key component faults of the equipment helps to quantify the fault risks, provides a basis for formulating maintenance strategies, enables the occurrence of equipment faults to be effectively predicted and intervened in a timely manner. Based on the equipment fault probability prediction data, analyzing the abnormal trend of the equipment operation state can reveal the abnormal fluctuation trends that occur during the equipment operation, and provide data support for adjusting the production plan in a timely manner and taking maintenance measures. Using the equipment abnormal trend analysis data to evaluate the fault risk level can classify the equipment faults of different risk levels, and thus achieve more accurate and targeted maintenance operations. Finally, integrating the equipment fault probability prediction data and the equipment fault risk level data can provide a comprehensive equipment fault prediction system, provide all-round data support for equipment maintenance decision-making, reduce the occurrence of sudden faults, optimize the equipment maintenance plan, and improve the overall efficiency and stability of the production system. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-restrictive embodiments read with reference to the accompanying drawings:

[0079] Figure 1 It is a schematic diagram of the modules of the online control system of the intelligent daily chemical production line of the present invention;

[0080] Figure 2 is Figure 1 a schematic diagram of the functional process of the equipment digital twin modeling module in

[0081] Figure 3 is Figure 1 a schematic diagram of the functional process of the process parameter optimization module in DETAILED DESCRIPTION OF THE EMBODIMENTS

[0082] The technical method of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0083] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0084] It should be understood that although terms such as "first" and "second" may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit can be called the second unit, and similarly the second unit can be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0085] To achieve the above object, please refer to Figures 1 to 3 , the present invention provides an on-line control system for an intelligent daily chemical production line, and the system includes the following modules:

[0086] An equipment digital twin modeling module, configured to obtain real-time operation state data of the equipment of the intelligent daily chemical production line, and perform digital twin modeling based on the real-time operation state data of the equipment to obtain an equipment digital twin model;

[0087] A process parameter optimization module, configured to perform spatio-temporal graph neural network simulation optimization calculation on each process parameter based on the equipment digital twin model, and perform multi-objective optimization simulation calculation to obtain equipment steady-state process parameter set data;

[0088] A modular production unit adjustment module, configured to optimize the modular production unit scheduling strategy based on the equipment steady-state process parameter set data, and perform dynamic characteristic modeling of the modular production unit to obtain optimized modular production unit data;

[0089] An automatic scheduling and Internet of Things transmission module, configured to perform automatic scheduling on the optimized modular production unit data, and perform Internet of Things remote data transmission on the automatic scheduling result to obtain equipment remote production scheduling instruction data;

[0090] The device monitoring and data analysis module is used to analyze the spatio-temporal evolution trend of the intelligent daily chemical production line based on the device remote production scheduling instruction data, and conduct real-time device monitoring to generate device operation analysis data;

[0091] The fault prediction and remote operation and maintenance module is used to identify the spatio-temporal pattern of the device health status based on the device operation analysis data, use the spatio-temporal pattern recognition result to predict the device fault and design the predictive maintenance plan, and conduct remote operation and maintenance on the intelligent daily chemical production line based on the predictive maintenance plan design result to obtain an optimized production control system.

[0092] In the embodiment of the present invention, refer to Figure 1 As shown, it is a schematic diagram of the step flow of an online control system for an intelligent daily chemical production line of the present invention. In this example, the online control system for the intelligent daily chemical production line includes the following modules:

[0093] S1: The device digital twin modeling module is used to obtain the real-time operation status data of the devices on the intelligent daily chemical production line, and conduct digital twin modeling based on the real-time operation status data of the devices to obtain a device digital twin model;

[0094] S2: The process parameter optimization module is used to conduct spatio-temporal graph neural network simulation optimization calculation on each process parameter based on the device digital twin model, and conduct multi-objective optimization simulation calculation to obtain the data of the device stable-state process parameter set;

[0095] S3: The modular production unit adjustment module is used to optimize the modular production unit scheduling strategy based on the data of the device stable-state process parameter set, and conduct dynamic characteristic modeling of the modular production unit to obtain optimized modular production unit data;

[0096] S4: The automatic scheduling and Internet of Things transmission module is used to automatically schedule the optimized modular production unit data, and conduct Internet of Things remote data transmission on the automatic scheduling result to obtain device remote production scheduling instruction data;

[0097] S5: The device monitoring and data analysis module is used to analyze the spatio-temporal evolution trend of the intelligent daily chemical production line based on the device remote production scheduling instruction data, and conduct real-time device monitoring to generate device operation analysis data;

[0098] S6: The fault prediction and remote operation and maintenance module is used to identify the spatio-temporal pattern of the device health status based on the device operation analysis data, use the spatio-temporal pattern recognition result to predict the device fault and design the predictive maintenance plan, and conduct remote operation and maintenance on the intelligent daily chemical production line based on the predictive maintenance plan design result to obtain an optimized production control system.

[0099] In the embodiments of the present invention, real-time operation status data of a device is obtained through a sensor array and a data acquisition board with a fixed sampling period. The signal is denoised by a Butterworth low-pass filter with an order of 4 in a digital filtering algorithm, and the operation status of each part of the device is accurately calculated by using numerical integration combined with the finite element analysis method, so as to generate a digital twin model of the device. According to the digital twin model of the device, a graph neural network with a fixed topological structure is constructed for each process parameter in the time and space dimensions. The data of each node is represented by a preset vector, and the relationship between nodes is characterized by a predefined correlation coefficient. Then, iterative operations are performed through a multi-objective constraint function and a gradient descent algorithm to output data of a stable-state process parameter set of the device. After receiving the data of the stable-state process parameter set of the device, the discrete event scheduling algorithm is used to strictly plan the operation time sequence of the production unit, and the ARIMA time series analysis is used to quantitatively model the dynamic characteristics of the modular production unit, so as to generate optimized modular production unit data. The optimized modular production unit data is automatically scheduled based on the state machine algorithm, and the data transmission is completed through a dedicated encryption channel by using fixed encryption parameters and an industrial-grade MQTT communication protocol to form device remote production scheduling instruction data. Taking the device remote production scheduling instruction data as input, the local regression algorithm and the real-time data acquisition device are used to continuously analyze the spatio-temporal evolution trend of the production line, and the operation status of the device is monitored according to a fixed data fusion method to generate device operation analysis data. According to the device operation analysis data, the Fourier transform combined with the clustering algorithm is used to strictly identify the spatio-temporal pattern of the device health status, and the decision tree rule is used to design the device fault prediction and predictive maintenance plan for the identification result. Subsequently, remote operation and maintenance operations are realized through a centralized control terminal, an industrial Ethernet, and the IEC standard protocol, so as to construct an optimized production control system. All modules are tightly connected in terms of data format and transmission rate through fixed parameters, preset algorithms, and standard communication protocols.

[0100] The online control system of the intelligent daily chemical production line proposed in the present invention integrates equipment digital twin modeling, process parameter optimization, modular production unit adjustment, automated scheduling and Internet of Things data transmission, equipment monitoring and data analysis, as well as fault prediction and remote operation and maintenance modules, which significantly improves the overall dynamic response capability of the production line and the accuracy of process parameter optimization. It uses real-time collected equipment operation status data to build a high-fidelity digital twin to achieve accurate reproduction of equipment status information, and determines process parameters based on detailed spatiotemporal simulation calculations. It generates a steady-state process parameter set through spatiotemporal graph neural network and multi-objective optimization algorithm, thereby effectively avoiding the lag and inefficiency of traditional empirical settings. At the same time, modular production unit adjustment is implemented. The scheduling strategy optimization and dynamic characteristic modeling based on real-time data can meet the needs of flexible production. The automated scheduling and Internet of Things data transmission module ensures that the scheduling instructions are issued immediately through fixed communication protocols and real-time data transmission mechanisms. The equipment monitoring and data analysis module ensures real-time feedback of operating status information through rigorous spatiotemporal evolution trend analysis and continuous monitoring. The fault prediction and remote operation and maintenance module relies on strict spatiotemporal pattern recognition and predictive maintenance solution design to achieve early warning of equipment failures and remote operation and maintenance operations, thereby reducing equipment failure rate and maintenance costs, ensuring production continuity, and realizing overall refined management and dynamic optimization of online control of the production line, significantly improving production efficiency and system operation stability.

[0101] Preferably, the device digital twin modeling module includes the following functions:

[0102] Obtain the real-time operation status data of the equipment of the intelligent daily chemical production line, and perform high-precision modeling of the geometric and physical characteristics of the equipment based on the real-time operation status data of the equipment to obtain the model data of the physical characteristics of the equipment;

[0103] Perform equipment dynamic behavior simulation based on equipment physical characteristic model data to obtain equipment dynamic behavior data, wherein the equipment dynamic behavior simulation includes vibration behavior simulation, thermodynamic behavior simulation, electromagnetic behavior simulation and fluid dynamic behavior simulation;

[0104] The device dynamic behavior data is used to perform an implicit representation method of neural radiation field to optimize the virtual mapping of device operation parameters and obtain device virtual mapping data;

[0105] Conduct electromagnetic field effect, material fatigue analysis and fluid interaction analysis on the virtual mapping data of the equipment, and build a multi-physics field coupling digital twin environment based on the results of electromagnetic field effect, material fatigue analysis and fluid interaction analysis to obtain the equipment digital twin environment data;

[0106] Based on the equipment digital twin environment data, the real-time operation status data of the equipment is integrated and analyzed to obtain a preliminary equipment digital twin model;

[0107] Anomaly detection is performed on the preliminary device digital twin model, and based on the anomaly detection results, the model parameters are optimized to obtain the device digital twin model.

[0108] As an embodiment of the present invention, refer to Figure 2 As shown in Figure 1 the schematic diagram of the functional flow of the device digital twin modeling module in

[0109] S11: Obtain the real-time operation status data of the devices on the intelligent daily chemical production line, and perform high-precision modeling of the geometric and physical characteristics of the devices based on the real-time operation status data of the devices to obtain the device physical characteristic model data;

[0110] S12: Perform device dynamic behavior simulation based on the device physical characteristic model data to obtain device dynamic behavior data, where the device dynamic behavior simulation includes vibration behavior simulation, thermodynamics behavior simulation, electromagnetic behavior simulation, and fluid dynamics behavior simulation;

[0111] S13: Use the device dynamic behavior data to optimize the virtual mapping of the device operation parameters by the neural radiance field implicit representation method to obtain the device virtual mapping data;

[0112] S14: Perform electromagnetic field effect, material fatigue analysis, and fluid interaction analysis on the device virtual mapping data, and construct a multi-physical field coupling digital twin environment based on the results of the electromagnetic field effect, material fatigue analysis, and fluid interaction analysis to obtain the device digital twin environment data;

[0113] S15: Perform fusion analysis on the real-time operation status data of the devices based on the device digital twin environment data to obtain the preliminary device digital twin model;

[0114] S16: Perform anomaly detection on the preliminary device digital twin model, and optimize the model parameters based on the anomaly detection results to obtain the device digital twin model.

[0115] In the embodiments of the present invention, a sensing network composed of an acceleration sensor with a fixed sampling frequency of 500 Hz, a thermocouple, and a strain gauge is utilized. Real-time operation state data of the device is obtained through a data acquisition circuit board, and the collected data is subjected to fourth-order Butterworth low-pass filtering and fixed-point downsampling processing to eliminate noise and suppress interference. Subsequently, laser scanning technology is used to obtain point cloud data, and parametric computer-aided design geometric algorithms and the finite element method are employed to extract the geometric dimensions of the device and determine the physical properties of the material. The stress distribution is calculated according to a grid size of 1 mm and a predetermined material parameter table to generate device physical characteristic model data; based on this device physical characteristic model data, dynamic behavior simulation operations are performed. Among them, for the vibration behavior simulation, explicit dynamic integration is carried out using the modal analysis algorithm, with a time step set to 0.001 s and a simulation duration of 10 s to solve for the natural frequencies and vibration modes of the structure. For the thermodynamics behavior simulation, the finite volume method is used to solve the heat conduction equation, with a spatial grid resolution set to 0.5 mm. For the electromagnetic behavior simulation, the discretized Maxwell's equations are used to calculate the electric and magnetic field distributions, and the potential is fixed in the boundary conditions. For the hydrodynamic behavior simulation, the computational fluid dynamics method is used to solve the Navier-Stokes equation, with the grid cell size set to 0.2 mm, thereby obtaining device dynamic behavior data; using the aforementioned device dynamic behavior data, the neural radiance field implicit representation method is adopted to optimize the virtual mapping of device operation parameters. Among them, a multi-layer perceptron containing three hidden layers with 64 neurons in each layer is constructed, and backpropagation is performed within 100 training cycles through a fixed learning rate of 0.01 and the mean square error loss function to adjust the weight parameters of each layer to generate device virtual mapping data; for the device virtual mapping data, first, the finite-difference time-domain algorithm is used to solve the electromagnetic field effect on a grid with a resolution of 0.1 mm, second, material fatigue analysis is performed through the stress-life curve and a fixed cyclic threshold calculation, and then the Euler-Lagrange coupling method is used to calculate the interaction between the fluid and the structure at a synchronous time step of 0.005 s. The three results are coupled with fixed interface variables to form multi-physics field coupling digital twin environment data; based on this device digital twin environment data, through Kalman filtering method (fixed covariance matrix) and weighted least squares method (preset weight) data fusion operations with the real-time operation state data, a preliminary device digital twin model is generated, and this model aims to reflect the current operation state of the device; for the preliminary device digital twin model, the statistical process control method is used to calculate the standard deviation of parameters within a fixed sample window (60 s) to determine the control limit, and the hypothesis testing method is used to determine the deviation threshold and then perform anomaly detection. Subsequently, the gradient descent algorithm with a fixed step size of 0.001 and a maximum of 50 iteration times is used for parameter optimization operations, and finally, a device digital twin model is generated.

[0116] The present invention can ensure the accuracy of the physical property model of the equipment and improve the ability of the model to reflect the true operating state of the equipment by obtaining the real-time operating state data of the intelligent daily chemical production line equipment and performing high-precision modeling of the geometric and physical properties of the equipment based on this. Based on this physical property model, dynamic behavior simulation of the equipment is carried out, so that the vibration, thermodynamics, electromagnetics, and fluid dynamics characteristics of the equipment are comprehensively simulated, thereby accurately predicting the dynamic performance of the equipment under different working conditions and improving the safety and stability of the equipment operation. The implicit representation method of the neural radiance field is optimized by using the equipment dynamic behavior data to make the virtual mapping of the equipment operation parameters more accurate, providing an efficient basis for parameter adjustment for equipment optimization control. By analyzing the electromagnetic field effect, material fatigue, and fluid interaction of the virtual mapping data, a digital twin environment with multi-physical field coupling is constructed, making the prediction of the equipment operating state more comprehensive and the optimization scheme more targeted. On this basis, the real-time operating state data of the equipment is fused and analyzed to establish an accurate digital twin model of the equipment, thereby improving the comprehensive perception ability of the equipment operating state. Further, anomaly detection is performed on the digital twin model and the model parameters are optimized, which can enhance the self-adaptability and anomaly recognition ability of the model, improve the accuracy of equipment predictive maintenance, reduce the equipment failure rate, extend the service life of the equipment, and ultimately improve the intelligent management level and production efficiency of the production line.

[0117] Preferably, the optimization of the virtual mapping of the equipment operation parameters by using the equipment dynamic behavior data for the implicit representation method of the neural radiance field includes:

[0118] Performing multi-dimensional feature extraction of the equipment operating state based on the equipment dynamic behavior data to obtain equipment dynamic behavior feature data;

[0119] Performing implicit representation encoding on the equipment dynamic behavior feature data and constructing a preliminary neural radiance field model based on the implicit representation encoding result to obtain initial equipment implicit representation model data;

[0120] Performing adaptive weight allocation and regularization strategy adjustment based on the initial equipment implicit representation model data to obtain structurally optimized equipment implicit representation model data;

[0121] Performing backpropagation training on the structurally optimized implicit representation model data and performing cross-validation and model parameter adjustment based on the backpropagation training result to obtain finely optimized equipment implicit representation model data;

[0122] Performing virtual mapping of the equipment operation parameters based on the finely optimized equipment implicit representation model data and performing error feedback correction on the virtual mapping result of the equipment operation parameters to obtain equipment virtual mapping data.

[0123] In the embodiments of the present invention, first, the device dynamic behavior data collected by a high-precision acceleration sensor, a temperature sensor, and a strain gauge are processed using the discrete wavelet transform algorithm and the fast Fourier transform algorithm to extract time-domain features (data mean, data variance, data kurtosis, data skewness), frequency-domain features (main frequency, amplitude spectrum, spectral energy distribution), and statistical features (minimum value, maximum value, standard deviation). Feature calculations are performed according to a fixed sampling window of 10 seconds and a data resampling interval of 0.1 second, and device dynamic behavior feature data is output; subsequently, the above feature data is input into an implicit representation coding unit constructed as a three-layer fully connected network, where the first and second layers are each configured with 64 nodes, and the third layer is configured with 32 nodes. The initial weights of all nodes are determined by random numbers uniformly distributed within the interval [-0.05, 0.05]. The activation function for all is the rectified linear unit, and the output layer uses a linear function for mapping. The implicit coding result is directly associated with the actual operating parameters of the device according to a fixed mapping relationship to form a preliminary neural radiance field construction scheme, and the initial device implicit representation model data is output; based on the initial device implicit representation model data, the adaptive weight distribution coefficient is calculated and determined using the softmax function for the output of each layer of nodes in the network, and an L2 regularization term is added to the loss function. The regularization coefficient is fixed at 0.001, and at the same time, the gradient descent algorithm is used for parameter update. The fixed learning rate is 0.005, and the momentum coefficient is set to 0.9. After a single iteration, the adaptive weight distribution and regularization strategy adjustment are completed, and the structure-optimized device implicit representation model data is generated; for the structure-optimized device implicit representation model data, the backpropagation algorithm with a batch size of 32 and 200 training epochs is used, and the mean squared error is used as the loss function. After each round of training, the 5-fold cross-validation method is used to verify the network output. After calculating the validation loss for each fold, the parameter adjustment amplitude is determined according to a fixed threshold of 1×10 -4 to update the network weights by recalculating the average value of the gradients for each fold, and finally, the finely optimized device implicit representation model data is output; based on the finely optimized device implicit representation model data, a virtual mapping of the device operating parameters is constructed using a fixed linear mapping function, whose mapping coefficient is determined by the production line calibration data. The implicit representation data is converted into device operating parameters, and by calculating the residual between the mapping result and the real-time collected parameters, the proportional coefficient in the proportional-integral control algorithm is set to 0.1 and the integral coefficient is set to 0.05 to perform error feedback correction on the mapping function. After multiple iterative adjustments, the device virtual mapping data is output.

[0124] By performing multi-dimensional feature extraction on the dynamic behavior data of the device, the core characteristics of the device operation status can be fully explored, and the recognition and modeling accuracy of the complex dynamic behavior of the device can be improved. Using the implicit representation coding method, the high-dimensional dynamic behavior feature data can be compressed into a low-dimensional latent space, reducing data redundancy while enhancing the information expression ability, thereby providing more compact and efficient input data for the construction of the neural radiance field model. Based on the initial implicit representation model data, adaptive weight allocation and regularization strategy adjustment are helpful to improve the adaptability of the model to different features, suppress the overfitting problem, and improve the generalization ability and stability of the model. By optimizing the implicit representation model through backpropagation training and combining cross-validation and parameter adjustment, the convergence and accuracy of the model can be further enhanced, ensuring that the mapping of the device operation status is more accurate. Finally, in the process of device virtual mapping, combined with the error feedback correction mechanism, the mapping result is closer to the real physical characteristics, improving the prediction accuracy and adaptive optimization ability of the model, providing more reliable data support for the intelligent optimization control of the device, thereby improving the device operation efficiency, reducing energy consumption, and enhancing the intelligent level of the production line.

[0125] Preferably, the process parameter optimization module includes the following functions:

[0126] Extract spatio-temporal dimension features from the device digital twin model to obtain process parameter spatio-temporal feature vector data;

[0127] Construct a spatio-temporal graph structure based on the process parameter spatio-temporal feature vector data to obtain dynamic spatio-temporal relationship graph data;

[0128] Perform spatio-temporal graph convolutional network modeling on the dynamic spatio-temporal relationship graph data to obtain a spatio-temporal feature fusion model;

[0129] Use the spatio-temporal feature fusion model to calculate multi-scale spatio-temporal attention to obtain dynamic weight allocation matrix data;

[0130] Based on the dynamic weight allocation matrix data, perform process parameter spatio-temporal evolution simulation to obtain parameter spatio-temporal response surface data;

[0131] Based on the parameter spatio-temporal response surface data, perform multi-objective genetic algorithm optimization and perform dynamic game theory conflict resolution on the multi-objective genetic algorithm optimization results to obtain device steady-state process parameter set data.

[0132] As an embodiment of the present invention, refer to Figure 3 shown as Figure 1 the functional flow schematic diagram of the process parameter optimization module in

[0133] S21: Extract the spatio-temporal dimensional features of the device digital twin model to obtain the spatio-temporal feature vector data of process parameters;

[0134] S22: Construct a spatio-temporal graph structure based on the spatio-temporal feature vector data of process parameters to obtain the dynamic spatio-temporal relationship graph data;

[0135] S23: Perform spatio-temporal graph convolutional network modeling on the dynamic spatio-temporal relationship graph data to obtain a spatio-temporal feature fusion model;

[0136] S24: Use the spatio-temporal feature fusion model to calculate the multi-scale spatio-temporal attention to obtain the dynamic weight allocation matrix data;

[0137] S25: Perform spatio-temporal evolution simulation of process parameters based on the dynamic weight allocation matrix data to obtain the spatio-temporal response surface data of parameters;

[0138] S26: Optimize based on the spatio-temporal response surface data of parameters using a multi-objective genetic algorithm, and perform dynamic game theory conflict resolution on the optimization results of the multi-objective genetic algorithm to obtain the data of the stable-state process parameter set of the device.

[0139] In the embodiment of the present invention, first, the discrete wavelet transform is used to decompose the device operation data collected in the device digital twin model according to a fixed time window of 10 seconds, and the two-dimensional Fourier transform is used to calculate the frequency domain change of the process parameters within a fixed resolution of 0.1 meter at each spatial position. The change trend of the parameters within each spatio-temporal segment is calculated by the least squares method, so as to extract the spatio-temporal feature vector data of the process parameters; then, the fixed threshold Euclidean distance is combined with the K-nearest neighbor algorithm (the K value is set to 5) to determine the adjacent connection relationship of each feature vector, and a spatio-temporal graph structure is constructed with the node attributes composed of the feature vector components and the edge weights generated according to the calculation results of the Euclidean distance, and the dynamic spatio-temporal relationship graph data is output; subsequently, graph convolution operations are performed on the dynamic spatio-temporal relationship graph data, a three-layer graph convolution structure is constructed, the convolution kernel size of each layer is fixed at 3, and the ReLU activation function is introduced in the convolution operation to update the node attributes and the adjacency matrix layer by layer to obtain the spatio-temporal feature fusion model; on this basis, the multi-head attention mechanism is implemented on the outputs of each node in the spatio-temporal feature fusion model, the number of multi-heads is fixed at 4 and the dimension of each head is fixed at 16, and the attention weight of each node at different scales is calculated and normalized by softmax to generate the dynamic weight allocation matrix data; using the dynamic weight allocation matrix data, the Crank-Nicolson difference scheme is used to discretize and solve the spatio-temporal evolution equation of the process parameters on the grid composed of a fixed spatial step of 0.1 meter and a time step of 0.05 second, and the process parameter values in each grid cell are updated hour by hour to generate the parameter spatio-temporal response surface data; finally, the multi-objective genetic algorithm optimization operation is performed on the parameter spatio-temporal response surface data, the population size is fixed at 50, the crossover rate is set to 0.8, the mutation rate is set to 0.05, the objective function is composed of the weighted sum of the process parameter stability and the energy consumption index, and after the Pareto optimal solution set is output by iterative solution, the conflict resolution operation is performed on the conflict solutions according to the fixed priority weight by using the dynamic game theory algorithm based on the Nash equilibrium, and finally the device steady-state process parameter set data is output.

[0140] Through the extraction of spatio-temporal dimensional features from the digital twin model of the equipment, the present invention can accurately capture the variation laws of process parameters at different time and space scales, and improve the comprehensive perception ability of the process state. Constructing a spatio-temporal graph structure can effectively reveal the dynamic correlation between process parameters, enhance the analysis ability of complex process evolution, and thus provide high-quality data support for subsequent modeling. Using a spatio-temporal graph convolutional network for modeling enables the model to efficiently extract spatio-temporal relationship features and improve the understanding ability of the dynamic change trend of process parameters. Combining multi-scale spatio-temporal attention calculation can achieve the accurate identification of key influencing factors and dynamic weight allocation, enhance the self-adaptability of the model, and improve the intelligent perception ability of the process optimization process. Based on the dynamic weight allocation matrix, spatio-temporal evolution simulation of process parameters is carried out, so that the change trends of process parameters under different working conditions can be visually expressed, thereby improving the prediction accuracy of complex process. Finally, through the optimization of the multi-objective genetic algorithm combined with the conflict resolution of dynamic game theory, the competitive relationship between process parameters can be effectively balanced, the stable operation state of the equipment can be optimized, and the process parameters can reach the optimal configuration under various constraints, thereby improving production efficiency, reducing energy consumption, and enhancing the stability and intelligent level of the production line.

[0141] Preferably, the modular production unit adjustment module includes the following functions:

[0142] Based on the data of the stable-state process parameter set of the equipment, perform parameter matching for the operation parameters of the modular production unit to obtain the parameter data of the equipment production unit;

[0143] Based on the initial production unit parameter data of the equipment, perform optimization analysis on the collaborative operation characteristics of each modular production unit to obtain the collaborative optimization data of the equipment production unit;

[0144] Use the collaborative optimization data of the equipment production unit to perform adjustment calculation on the process flow parameters of each modular production unit to obtain the process parameter data of the optimized production unit of the equipment;

[0145] Based on the process parameter data of the optimized production unit of the equipment, perform dynamic adaptation adjustment on the production capacity of the modular production unit to obtain the data of the optimized modular production unit.

[0146] Based on the data of the equipment's steady-state process parameter set, the embodiments of the present invention first perform the matching of the operating parameters of the modular production units. Using the production data collected by the control system and combining with the process requirements of the equipment, linear regression and Kalman filtering algorithms are adopted to match the operating parameters of the production units, obtaining the parameter data of the equipment production units, which include the operating status and performance indicators of each module. Next, based on the initial production unit parameter data of the equipment, the collaborative operation characteristics optimization analysis of each modular production unit is carried out. Multiple regression analysis and principal component analysis (PCA) are used to evaluate the collaborative effects between different modules, identify the mutual influences of each module, optimize the cooperation relationship between each module, and obtain the collaborative optimization data of the equipment production units, which include the optimized collaborative operation characteristics of each modular production unit. Then, using the collaborative optimization data of the equipment production units, the adjustment calculation of the process flow parameters of each modular production unit is carried out. Genetic algorithm (GA) or particle swarm optimization (PSO) is used to adjust the process flow of each module. By adjusting the input and output flow rates and reaction temperature parameters, each module can achieve the optimal cooperation at different production stages, obtaining the process parameter data of the optimized production unit of the equipment, which include the new process parameter range and the adjusted production conditions. Finally, based on the process parameter data of the optimized production unit of the equipment, the dynamic adaptation adjustment of the production capacity of the modular production unit is carried out. Using dynamic programming and real-time feedback mechanism, combined with the actual load of the production line and the market demand forecast, the production capacity of each production unit is adaptively adjusted, obtaining the optimized modular production unit data.

[0147] By matching the operating parameters of the modular production units based on the data of the equipment's steady-state process parameter set, the present invention can ensure that each production unit accurately adjusts its operating parameters on the basis of the optimized process parameters, thereby improving production efficiency, reducing energy consumption, and ensuring the consistency and quality of products. By performing the collaborative operation characteristics optimization analysis of each modular production unit based on the initial production unit parameter data of the equipment, the collaborative effects between production units can be identified, providing an accurate optimization scheme for the collaborative work of multiple modules, and avoiding the situation where the improvement of a single production unit cannot bring about the improvement of the overall benefit. By using the collaborative optimization data of the equipment production units to carry out the adjustment calculation of the process flow parameters of each modular production unit, the process flows can be refined, ensuring that each module can operate efficiently under the collaborative effect of the entire production line, and improving production capacity and production stability. By performing the dynamic adaptation adjustment of the production capacity of the modular production unit based on the process parameter data of the optimized production unit of the equipment, different demand changes and production loads during the production process can be flexibly responded to, ensuring that the production line can quickly adapt and operate efficiently when facing different production demands, thereby maximizing production benefits and the resource utilization rate of the system.

[0148] Preferably, the automated scheduling and IoT transmission module includes the following functions:

[0149] Decompose the production tasks for the optimized modular production unit data to obtain the equipment production task decomposition data;

[0150] Perform production resource allocation calculations based on the equipment production task decomposition data to obtain the equipment production resource allocation data;

[0151] Use the equipment production resource allocation data to optimize and adjust the production scheduling plan for the intelligent daily chemical production line to obtain the equipment production scheduling optimization data;

[0152] Monitor and verify the scheduling execution status based on the equipment production scheduling optimization data to obtain the equipment automated scheduling result data;

[0153] Perform Internet of Things data transmission based on the equipment automated scheduling result data to generate remote instructions to obtain the equipment remote production scheduling instruction data.

[0154] In the embodiments of the present invention, the production tasks of the optimized modular production unit data are decomposed. First, the functional requirements and production capabilities of each modular production unit are used as inputs. By using the production task analysis tool, through analyzing task types, production steps, time requirements, and equipment resource requirements, the overall production goal is decomposed into tasks of individual processes, and priorities and time windows are assigned to each task, obtaining equipment production task decomposition data, which includes the specific descriptions and process parameters of each production task. Based on the equipment production task decomposition data, production resource allocation calculations are performed. Using resource allocation optimization algorithms (such as linear programming or integer programming), the equipment, raw materials, and human resources required for the tasks are reasonably allocated to ensure that each task can be completed within the specified time, and based on the availability of production line resources and the resource consumption of each task, equipment production resource allocation data is obtained, including the specific allocation ratios and usage times of each resource. The production scheduling plan of the intelligent daily chemical production line is optimized and adjusted using the equipment production resource allocation data. Combining the priorities of production tasks, equipment load conditions, and resource usage conditions, scheduling optimization algorithms (such as genetic algorithms or simulated annealing algorithms) are used to optimize and adjust the scheduling plan of the production line to ensure the optimality of the task execution order and resource allocation, obtaining equipment production scheduling optimization data, which includes the adjusted production order and resource configuration. Based on the equipment production scheduling optimization data, the monitoring and verification of the scheduling execution status are carried out. The execution of each task on the production line is monitored in real time, the operation data of each module is collected and compared with the preset scheduling plan to verify whether the tasks are proceeding smoothly as planned. Each production link is tracked through a status monitoring tool (such as a PLC control system or a real-time data collection system), obtaining equipment automated scheduling result data, which includes the deviation between the actual execution situation and the expected goal of each task. Based on the equipment automated scheduling result data, Internet of Things data transmission is performed to generate remote instructions. Using an Internet of Things platform (such as the MQTT or HTTP protocol), the real-time scheduling result data is transmitted to the remote monitoring system to generate instructions and transmit them to each relevant production unit, obtaining equipment remote production scheduling instruction data.

[0155] By decomposing production tasks based on the optimized modular production unit data, the present invention can refine complex production tasks into more operable units, making the production process clearer and more systematic, and enhancing the controllability and monitorability of the production process. Based on the production resource allocation calculation using the production task decomposition data of the equipment, while ensuring the rational allocation of resources, the utilization rate of resources can be maximized, thereby reducing resource waste and enhancing the overall efficiency of the production line. Using the production resource allocation data of the equipment to optimize and adjust the production scheduling plan of the intelligent daily chemical production line can flexibly adjust the production plan in real time according to the resource allocation situation, ensure the optimization of production scheduling and adapt to dynamic changes, and enhance the flexibility and response speed of the production line. Monitoring and verifying the scheduling execution status based on the optimized data of equipment production scheduling helps to track the execution of the production process in real time, discover and correct problems in a timely manner, and ensure the efficient execution of the scheduling plan. Transmitting Internet of Things data based on the automated scheduling result data of the equipment to generate remote instructions can realize the remote control and monitoring of the production line, enhance the flexibility and management efficiency of production, provide real-time response capabilities for enterprises in dealing with emergencies during the production process, and thus enhance the intelligence and automation level of the production line.

[0156] Preferably, the equipment monitoring and data analysis module includes the following functions:

[0157] Collect the operating status data of each modular production unit of the intelligent daily chemical production line based on the equipment remote production scheduling instruction data to obtain the equipment production unit operating status data;

[0158] Detect abnormal key operating parameters of the equipment production unit operating status data to obtain equipment abnormal detection data;

[0159] Analyze the health status trend of the key components of the equipment based on the equipment abnormal detection data to obtain the equipment health status analysis data;

[0160] Analyze the production efficiency of the equipment and evaluate its stability based on the equipment health status analysis data to obtain the equipment operation analysis data.

[0161] In the embodiments of the present invention, based on the device remote production scheduling instruction data, the operation status of each modular production unit of the intelligent daily chemical production line is collected. By using the embedded data acquisition module and the PLC control system, the operation status data of each production unit is obtained in real time to get the operation status data of the device production unit, ensuring that the collected data includes the specific operation status of each module. The key operation parameter anomaly detection is carried out on the operation status data of the device production unit. Combining with the set threshold standard, through the anomaly detection algorithm (such as the anomaly detection method based on mean and standard deviation), the key operation parameters of each production unit are monitored in real time to identify the abnormal parameters beyond the set range, obtaining the device anomaly detection data, including each detected abnormal event and its abnormal degree. Based on the device anomaly detection data, the health status trend analysis of the key components of the device is carried out. Using the time series analysis method, through the analysis of the historical operation data and the anomaly detection data, the health status trend curve of the key components of the device is drawn to evaluate the change of the health status of the key components, obtaining the device health status analysis data, including the health trend, risk level and potential time point of failure occurrence of each component. Based on the device health status analysis data, the production efficiency analysis and stability evaluation of the device are carried out. By adopting the production efficiency analysis model, through analyzing the production efficiency of the device in different health states, the operation stability of the device is evaluated, obtaining the device operation analysis data, including the fluctuation of the device efficiency, the unstable factors occurring in the production process and the impact on the overall production process.

[0162] In the present invention, by collecting the operation status of each modular production unit of the intelligent daily chemical production line based on the device remote production scheduling instruction data, the operation status data of each module of the production line can be obtained in real time, ensuring that every link in the production process can be accurately monitored, which helps to quickly identify potential problems. The key operation parameter anomaly detection of the operation status data of the device production unit can detect the abnormal fluctuations and potential faults in the device operation in real time, prevent the problems from deteriorating further, thereby reducing the downtime and maintenance costs. Based on the device anomaly detection data, the health status trend analysis of the key components of the device is carried out, which can accurately evaluate the health status of the key components of the device, predict the future health trend of the device, ensure that timely measures are taken for maintenance and repair, and avoid the risk of production interruption caused by sudden failures. Based on the device health status analysis data, the production efficiency analysis and stability evaluation of the device are carried out, which can comprehensively evaluate the production efficiency and operation stability of the device, help the management personnel optimize the production scheduling, adjust the operation strategy, improve the overall performance of the device, ensure that the production line remains efficient and stable during long-term operation, thereby maximizing the production benefits and reducing the impact of production fluctuations.

[0163] Preferably, the production efficiency analysis and stability evaluation of the device based on the device health status analysis data include:

[0164] Perform time series analysis on the device health status analysis data to obtain device operation time characteristic data;

[0165] Calculate the production task completion rate based on the device operation time characteristic data to obtain device production task completion rate data;

[0166] Build a device spatio-temporal key feature model based on the device production task completion rate data to obtain device energy consumption analysis data;

[0167] Detect the fluctuations in the device operation stability based on the device energy consumption analysis data to obtain device stability detection data;

[0168] Evaluate the load status of the device's key components based on the device stability detection data to obtain device key component load data;

[0169] Calculate the device production efficiency using the device production task completion rate data and the device key component load data to obtain device production efficiency analysis data;

[0170] Conduct spatio-temporal trend analysis on the long-term operation stability of the device based on the device production efficiency analysis data to obtain device operation analysis data.

[0171] Embodiments of the present invention are based on the equipment health status analysis data of an intelligent daily chemical production line, use time series decomposition algorithms to perform trend analysis on the data, extract periodic change characteristics, and calculate the operating time characteristics of the equipment in combination with the exponential smoothing method to obtain equipment operating time characteristic data; based on the equipment operating time characteristic data, according to the production task scheduling records, compare the actual operating time of the equipment with the planned production time, and use the weighted task completion rate calculation method to calculate the task execution efficiency of each production unit to obtain equipment production task completion rate data; based on the equipment production task completion rate data, adopt a spatio-temporal data correlation analysis method to model the key spatio-temporal characteristics during the equipment operation process, extract the energy consumption characteristic parameters of the equipment in different time periods, and calculate the energy consumption status of each production unit in combination with the energy flow analysis method to obtain equipment energy consumption analysis data; based on the equipment energy consumption analysis data, use the stability fluctuation detection algorithm to identify the unstable factors existing in the equipment operation process by calculating the power change rate, load fluctuation frequency, and energy consumption peak distribution to obtain equipment stability detection data; based on the equipment stability detection data, for the historical load conditions of the key components of the equipment, adopt the stress-strain curve analysis method to evaluate the long-term stress and fatigue of the key components of the equipment, calculate the load level and bearing capacity of the key components to obtain equipment key component load data; use the equipment production task completion rate data and the equipment key component load data, adopt the production efficiency calculation formula, combine the unit energy consumption output ratio, task completion efficiency, and equipment utilization rate to calculate the overall production efficiency of the equipment to obtain equipment production efficiency analysis data; based on the equipment production efficiency analysis data, use the spatio-temporal trend analysis method to analyze the change of the production efficiency of the equipment in different time periods by constructing an operating state data curve with a long time span, and combine the load balance analysis to predict the stability trend of the long-term operation of the equipment to obtain equipment operation analysis data.

[0172] Through time series analysis of the equipment health status analysis data, the present invention can accurately extract the long-term operation characteristics of the equipment, identify the time characteristics of different stages in the equipment operation process, so as to provide accurate data support for the subsequent calculation of the production task completion rate. In the process of calculating the production task completion rate, combining the time characteristic data can accurately measure the execution efficiency of the production unit, improve the matching degree of the production plan, and optimize the production scheduling strategy. Based on the production task completion rate data, the spatio-temporal key characteristics of the equipment are modeled, so that the energy consumption distribution of the equipment can be accurately characterized in the time and space dimensions, thereby revealing the energy consumption characteristics and energy efficiency bottlenecks of each production unit, and providing a basis for energy-saving optimization. On the basis of energy consumption analysis, the equipment operation stability fluctuation detection is carried out, which can accurately identify the stability change of the equipment under different working conditions, timely detect abnormal operation fluctuations, and prevent the equipment from unplanned shutdown caused by abnormal energy consumption or load fluctuations. By evaluating the load status of the key components based on the equipment stability detection data, the bearing capacity of the key components of the equipment can be evaluated in real time, and then the load distribution can be optimized, the service life of the key components of the equipment can be extended, and the overall reliability of the production line can be improved. By combining the production task completion rate data and the key component load data to calculate the production efficiency of the equipment, the resource utilization rate, energy efficiency ratio and equipment utilization rate in the production process can be comprehensively analyzed, providing a reliable basis for optimizing the production efficiency. Based on the equipment production efficiency analysis data, the spatio-temporal trend analysis of the long-term operation stability is carried out, which can reveal the trend changes of the equipment in the long-term operation, predict the fluctuation trend of the production efficiency in advance, and ensure the efficient and stable operation of the production line under different production loads and environmental change conditions.

[0173] Preferably, the fault prediction and remote operation and maintenance module includes the following functions:

[0174] Based on the equipment operation analysis data, a historical operation record trend model is established to obtain equipment operation trend data;

[0175] Based on the equipment operation trend data, a historical fault characteristic pattern matching analysis of the key components of the equipment is carried out to obtain equipment fault prediction data;

[0176] Based on the equipment fault prediction data, an optimization calculation of the maintenance strategy is carried out to obtain equipment predictive maintenance plan data;

[0177] Based on the equipment predictive maintenance plan data, remote operation and maintenance of the intelligent daily chemical production line is carried out to obtain an optimized production control system.

[0178] In the embodiments of the present invention, historical operation record trend modeling is performed based on device operation analysis data. First, historical data is extracted from various operation parameters and status data of the device. The time series analysis method is used to model the operation data of the device, identify the long-term operation mode of the device, predict the change trend of various indicators of the device in a future period of time, obtain device operation trend data, and reflect the change of the operation state of the device in different time periods. Based on the device operation trend data, historical fault feature pattern matching analysis of key components of the device is carried out. By extracting the historical fault records of key components of the device and using fault pattern recognition algorithms (such as K-nearest neighbor algorithm, support vector machine) for feature matching, the similarity between the current operation trend and the historical fault pattern is compared to determine the type of fault that occurs in the device, and device fault prediction data is obtained to help identify device problems in advance. Based on the device fault prediction data, maintenance strategy optimization calculation is performed. By combining the device fault prediction data with the maintenance record data and using optimization algorithms (such as genetic algorithm, particle swarm optimization) to optimize the maintenance strategy calculation, the most appropriate maintenance time, maintenance items, and maintenance resource allocation are determined, and device predictive maintenance plan data is obtained to ensure that necessary repairs and maintenance are carried out before the device fails. Based on the device predictive maintenance plan data, remote operation and maintenance of the intelligent daily chemical production line are carried out. Through the remote monitoring system, according to the optimized maintenance plan, maintenance resources are scheduled, and maintenance personnel are arranged to perform remote operations or on-site services to promptly handle device failures, and an optimized production control system is obtained.

[0179] Through historical operation record trend modeling based on device operation analysis data, the present invention can identify potential problems in the device operation process, reveal the stability and performance change trend of the device in long-term operation, and provide data support for subsequent fault prediction and maintenance decision-making. Based on the device operation trend data, historical fault feature pattern matching analysis of key components of the device can identify the fault pattern that occurs in the device in advance, and predict the future fault risk by comparing historical data, thus avoiding production interruption caused by sudden failures. Based on the device fault prediction data, maintenance strategy optimization calculation can formulate a scientific and reasonable predictive maintenance plan, ensure that the device is repaired in time before the fault occurs, reduce the occurrence of sudden failures, and improve the reliability of the device and the continuous operation ability of the production line. By performing remote operation and maintenance of the intelligent daily chemical production line based on the device predictive maintenance plan data, the health status of the device can be grasped in real time, the maintenance plan can be adjusted in time, the stability and operation efficiency of the production system are improved, and finally the optimization of the production control system is realized, enabling the device to operate in an efficient and stable state, reducing the maintenance cost and downtime, and ensuring the smooth progress of production tasks.

[0180] Preferably, the historical fault feature pattern matching analysis of key components of the device based on the device operation trend data includes:

[0181] Extract historical fault records from the device operation trend data and classify the fault modes to obtain the device fault mode classification data;

[0182] Based on the device fault mode classification data, conduct statistical analysis of the characteristic parameters of each fault mode to obtain the device fault characteristic parameter data;

[0183] Based on the device fault characteristic parameter data, perform characteristic matching calculations for the operating status of the device's key components to obtain the device fault characteristic matching data;

[0184] Use the device fault characteristic matching data to calculate the probability of occurrence of faults in the device's key components to obtain the device fault probability prediction data;

[0185] Based on the device fault probability prediction data, conduct an analysis of the abnormal trend of the device operation status to obtain the device abnormal trend analysis data;

[0186] Use the device abnormal trend analysis data to evaluate the fault risk level to obtain the device fault risk level data;

[0187] Integrate the device fault probability prediction data and the device fault risk level data to obtain the device fault prediction data.

[0188] In the embodiments of the present invention, historical fault records are extracted from the device operation trend data. First, fault records are extracted from the historical operation data of the device. By querying the fault logs and maintenance records of the device, historical fault events are identified, and these events are classified according to the fault types to obtain device fault mode classification data. The classification basis includes the frequency of fault occurrence, the time period of occurrence, the faulty components and their damage characteristics. Based on the device fault mode classification data, statistical analysis of the characteristic parameters of each fault mode is carried out. Statistical methods are used to extract the key characteristics of each fault mode, and the relevant characteristic parameters are calculated to obtain device fault characteristic parameter data. Based on the device fault characteristic parameter data, characteristic matching calculation of the operation state of the device's key components is carried out. The matching algorithm (such as similarity calculation or dynamic time warping) is used to compare the current operation data of the device with the historical fault characteristics, and the states similar to the historical fault modes are identified to obtain device fault characteristic matching data, which reflects the matching degree between the current operation state of the device's key components and the historical fault modes. The probability of fault occurrence of the device's key components is calculated using the device fault characteristic matching data. Combining the matching data and the statistical analysis results, a probability calculation model (such as Bayesian inference or Poisson distribution) is used to calculate the probability of fault occurrence of each key component to obtain device fault probability prediction data, which provides a basis for subsequent maintenance decisions. Based on the device fault probability prediction data, analysis of the abnormal trend of the device operation state is carried out. Trend analysis methods (such as linear regression or time series analysis) are used, combined with the device fault prediction data, to identify the abnormal change trend of the device operation state to obtain device abnormal trend analysis data, and predict the future operation risk of the device. The device abnormal trend analysis data is used to evaluate the fault risk level. Risk assessment methods (such as fuzzy logic or weighted scoring method) are used to evaluate the various risks of the device, and the possibility and severity of future fault occurrence of the device are obtained to obtain device fault risk level data, which reflects the risk level of the device operation. The device fault probability prediction data and the device fault risk level data are integrated. A data fusion method (such as weighted average or data synthesis) is used to combine the fault probability and risk level information to generate device fault prediction data.

[0189] By extracting historical fault records from the equipment operation trend data and classifying the fault modes, the present invention can deeply analyze the historical fault conditions of the equipment, identify the characteristics of different fault modes, and provide basic data for subsequent fault prediction. Based on the equipment fault mode classification data, statistical analysis of the characteristic parameters of each fault mode helps to reveal the specific manifestations and key features of different fault modes, helps to accurately evaluate the potential risks of equipment operation, and provides accurate parameter basis for subsequent fault diagnosis and prevention. Based on the equipment fault characteristic parameter data, calculating the matching of the operating state characteristics of the key components of the equipment can effectively identify the abnormal state of the key components in actual operation, detect fault signs in advance, and thus avoid equipment damage and production stagnation. Using the equipment fault characteristic matching data to calculate the occurrence probability of faults in the key components of the equipment helps to quantify the fault risks, provides a basis for formulating maintenance strategies, enables the occurrence of equipment faults to be effectively predicted and intervened in a timely manner. Based on the equipment fault probability prediction data, analyzing the abnormal trend of the equipment operation state can reveal the abnormal fluctuation trend that appears in the equipment operation process, and provide data support for adjusting the production plan in a timely manner and taking maintenance measures. Using the equipment abnormal trend analysis data to evaluate the fault risk level can classify and process equipment faults with different risk levels, so as to achieve more accurate and targeted maintenance operations. Finally, integrating the equipment fault probability prediction data and the equipment fault risk level data can provide a comprehensive equipment fault prediction system, provide all-round data support for equipment maintenance decision-making, reduce the occurrence of sudden faults, optimize the equipment maintenance plan, and improve the overall efficiency and stability of the production system.

[0190] Preferably, the analysis of the abnormal trend of the equipment operation state based on the equipment fault probability prediction data includes:

[0191] Performing time series decomposition on the equipment fault probability prediction data to obtain equipment fault probability time series data;

[0192] Based on the equipment fault probability time series data, fitting the trend of the operating parameters of the key components of the equipment to obtain equipment operating parameter trend data;

[0193] Detecting abnormal fluctuation points in the equipment operating parameter trend data to obtain equipment operation abnormal fluctuation data;

[0194] Classifying the equipment operation abnormal fluctuation data to obtain equipment abnormal fluctuation mode data;

[0195] Based on the equipment abnormal fluctuation mode data, comparing the historical abnormal trends of the key components to obtain equipment abnormal trend comparison data;

[0196] Using the equipment abnormal trend comparison data to perform data fault trend modeling to generate equipment abnormal trend analysis data.

[0197] In the embodiments of the present invention, time series decomposition is performed on the device failure probability prediction data. First, a seasonal decomposition method (such as STL decomposition) is used to separate the trend component, seasonal component, and residual component in the device failure probability prediction data, obtaining the device failure probability time series data, which reflects the change trend of the device failure occurrence probability. Based on the device failure probability time series data, trend fitting of the operating parameters of the device's key components is performed. The least squares method or other regression analysis methods are used to perform trend fitting on the operating parameters of the device's key components, obtaining the device operating parameter trend data, which reveals the long-term change trend of the device components. Detection of abnormal fluctuation points is performed on the device operating parameter trend data. Statistical analysis methods (such as Z-score or standard deviation analysis) are used to identify the fluctuation points in the device operating parameter trend data, identifying the abnormal fluctuations that significantly deviate from the normal trend within a certain time window, obtaining the device operating abnormal fluctuation data. Pattern classification is performed on the device operating abnormal fluctuation data. Clustering analysis methods (such as K-means or DBSCAN) are used to classify different types of abnormal fluctuations according to their characteristics, obtaining the device abnormal fluctuation pattern data, which identifies different fluctuation types and their occurrence rules. Based on the device abnormal fluctuation pattern data, comparison of the historical abnormal trends of the key components is performed. By comparing the abnormal fluctuation pattern of the current device with the abnormal patterns in the historical data, the comparison data of the historical abnormal trend and the current trend is obtained using a similarity calculation method (such as cosine similarity or dynamic time warping), revealing whether the device is in the evolution stage of an abnormal trend. Using the device abnormal trend comparison data, data failure trend modeling is performed. Combining the historical failure data and the current trend comparison data, a time series prediction model (such as ARIMA or LSTM neural network) is used to model the failure trend of the device, generating the device abnormal trend analysis data.

[0198] Through time series decomposition of the equipment failure probability prediction data, the present invention can extract the probability change trend of equipment failure into clear time series data, helping to identify potential periodic changes or sudden risks of equipment failure during operation. Trend fitting of the operating parameters of key components of the equipment based on the time series data of equipment failure probability helps to deeply analyze the working performance of key components of the equipment in different time periods, so as to predict the time period and cause of key component failure. Detection of abnormal fluctuation points in the trend data of equipment operating parameters can effectively detect abnormal fluctuations occurring during equipment operation, and early warning of failures can be given to avoid the expansion of failures or production interruption. By classifying the patterns of equipment operation abnormal fluctuation data, different types of abnormal fluctuation patterns can be accurately identified, and then the potential causes and influence ranges of different failures can be determined, providing a scientific basis for subsequent fault diagnosis. By comparing the historical abnormal trends of key components based on the equipment abnormal fluctuation pattern data, the long-term performance and abnormal trends of key components of the equipment can be accurately identified by comparing historical data, providing a dynamic monitoring perspective for equipment health management. Finally, by using the equipment abnormal trend comparison data for data fault trend modeling to generate equipment abnormal trend analysis data, the possibility of future failures can be predicted through model-based analysis, and maintenance plans and countermeasures can be formulated in advance, so as to effectively reduce the risk of production interruption and improve the reliability and efficiency of equipment operation.

[0199] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is not limited by the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be included in the present invention.

[0200] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent daily chemical production line online control system, characterized in that: Includes the following modules: The equipment digital twin modeling module is used to obtain the real-time operation status data of the equipment of the intelligent daily chemical production line, and perform digital twin modeling based on the real-time operation status data of the equipment to obtain the equipment digital twin model; The process parameter optimization module is used to perform spatiotemporal graph neural network simulation optimization calculations of various process parameters based on the equipment digital twin model, and perform multi-objective optimization simulation calculations to obtain the equipment steady-state process parameter set data; Modular production unit adjustment module, used to optimize the modular production unit scheduling strategy based on the equipment steady-state process parameter set data, and to model the dynamic characteristics of the modular production unit to obtain optimized modular production unit data; The automatic scheduling and IoT transmission module is used to automatically schedule the optimized modular production unit data and transmit the automatic scheduling results via IoT remote data to obtain the equipment remote production scheduling instruction data; The equipment monitoring and data analysis module is used to analyze the temporal and spatial evolution trend of the intelligent daily chemical production line based on the equipment remote production scheduling instruction data, and to perform real-time equipment monitoring and generate equipment operation analysis data; The fault prediction and remote operation and maintenance module is used to perform spatiotemporal pattern recognition of equipment health status based on equipment operation analysis data, use the spatiotemporal pattern recognition results to predict equipment faults and design predictive maintenance plans, and remotely operate and maintain the intelligent daily chemical production line based on the predictive maintenance plan design results to obtain an optimized production control system.

2. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The device digital twin modeling module includes the following functions: Obtain the real-time operation status data of the equipment of the intelligent daily chemical production line, and perform high-precision modeling of the geometric and physical characteristics of the equipment based on the real-time operation status data of the equipment to obtain the model data of the physical characteristics of the equipment; Perform equipment dynamic behavior simulation based on equipment physical characteristic model data to obtain equipment dynamic behavior data, wherein the equipment dynamic behavior simulation includes vibration behavior simulation, thermodynamic behavior simulation, electromagnetic behavior simulation and fluid dynamic behavior simulation; The device dynamic behavior data is used to perform an implicit representation method of neural radiation field to optimize the virtual mapping of device operation parameters and obtain device virtual mapping data; Conduct electromagnetic field effect, material fatigue analysis and fluid interaction analysis on the virtual mapping data of the equipment, and build a multi-physics field coupling digital twin environment based on the results of electromagnetic field effect, material fatigue analysis and fluid interaction analysis to obtain the equipment digital twin environment data; Based on the equipment digital twin environment data, the real-time operation status data of the equipment is integrated and analyzed to obtain a preliminary equipment digital twin model; Anomaly detection is performed on the preliminary equipment digital twin model, and model parameters are optimized based on the anomaly detection results to obtain the equipment digital twin model.

3. The intelligent daily chemical production line online control system according to claim 2 is characterized in that: The method of optimizing the virtual mapping of equipment operation parameters by using the equipment dynamic behavior data to perform the neural radiation field implicit representation method comprises: Extract multi-dimensional features of equipment operation status based on equipment dynamic behavior data to obtain equipment dynamic behavior feature data; Implicitly represent and encode the dynamic behavior feature data of the device, and construct a preliminary neural radiation field model based on the implicit representation encoding results to obtain the initial device implicit representation model data; Based on the initial device implicit representation model data, adaptive weight allocation and regularization strategy adjustment are performed to obtain the structure optimized device implicit representation model data; Perform back propagation training on the implicit representation model data of the structural optimization, and perform cross-validation and model parameter adjustment based on the back propagation training results to obtain the implicit representation model data of the finely optimized device; The virtual mapping of equipment operation parameters is performed based on the finely optimized equipment implicit representation model data, and the error feedback correction is performed on the virtual mapping result of the equipment operation parameters to obtain the equipment virtual mapping data.

4. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The process parameter optimization module includes the following functions: Extract the spatiotemporal dimensional features of the equipment digital twin model to obtain the spatiotemporal feature vector data of the process parameters; Based on the spatiotemporal characteristic vector data of process parameters, the spatiotemporal graph structure is constructed to obtain dynamic spatiotemporal relationship graph data; The spatiotemporal graph convolutional network modeling is performed on the dynamic spatiotemporal relationship graph data to obtain the spatiotemporal feature fusion model; Use the spatiotemporal feature fusion model to perform multi-scale spatiotemporal attention calculations to obtain dynamic weight allocation matrix data; Based on the dynamic weight allocation matrix data, the spatiotemporal evolution of process parameters is simulated to obtain the spatiotemporal response surface data of parameters; Multi-objective genetic algorithm optimization is carried out based on parameter space-time response surface data, and the dynamic game theory conflict resolution is performed on the multi-objective genetic algorithm optimization results to obtain the steady-state process parameter set data of the equipment.

5. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The modular production unit adjustment module includes the following functions: Based on the equipment steady-state process parameter set data, modular production unit operation parameter matching is performed to obtain equipment production unit parameter data; Based on the initial production unit parameter data of the equipment, the coordinated operation characteristics of each modular production unit are optimized and analyzed to obtain the equipment production unit coordinated optimization data; Use the equipment production unit collaborative optimization data to calculate the process parameters of each modular production unit to adjust the process parameters of the equipment optimized production unit; Based on the equipment optimization production unit process parameter data, the modular production unit capacity is dynamically adapted and adjusted to obtain optimized modular production unit data.

6. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The automated scheduling and IoT transmission module includes the following functions: Decomposing the optimized modular production unit data into production tasks to obtain equipment production task decomposition data; Perform production resource allocation calculation based on equipment production task decomposition data to obtain equipment production resource allocation data; Use equipment production resource allocation data to optimize and adjust the production scheduling plan of the intelligent daily chemical production line to obtain equipment production scheduling optimization data; Monitor and verify the scheduling execution status based on the equipment production scheduling optimization data to obtain the equipment automation scheduling result data; Based on the equipment automation scheduling result data, IoT data transmission is performed to generate remote instructions to obtain equipment remote production scheduling instruction data.

7. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The equipment monitoring and data analysis module includes the following functions: Based on the equipment remote production scheduling instruction data, the operation status of each modular production unit of the intelligent daily chemical production line is collected to obtain the equipment production unit operation status data; Perform abnormality detection on key operating parameters of the equipment production unit operation status data to obtain equipment abnormality detection data; Perform health status trend analysis of key equipment components based on equipment anomaly detection data to obtain equipment health status analysis data; Based on the equipment health status analysis data, equipment production efficiency analysis and stability assessment are performed to obtain equipment operation analysis data.

8. The intelligent daily chemical production line online control system according to claim 7 is characterized in that: The equipment production efficiency analysis and stability assessment based on equipment health status analysis data includes: Perform time series analysis on equipment health status analysis data to obtain equipment operation time characteristic data; Calculate the production task completion rate based on the equipment operation time characteristic data to obtain the equipment production task completion rate data; Based on the equipment production task completion rate data, the equipment time-space key characteristics modeling is carried out to obtain the equipment energy consumption analysis data; Perform equipment operation stability fluctuation detection based on equipment energy consumption analysis data to obtain equipment stability detection data; Evaluate the load status of key components of equipment based on equipment stability detection data to obtain load data of key components of equipment; Calculate equipment production efficiency using equipment production task completion rate data and equipment key component load data to obtain equipment production efficiency analysis data; Based on the equipment production efficiency analysis data, the spatiotemporal trend analysis of the equipment's long-term operation stability is performed to obtain the equipment operation analysis data.

9. The intelligent daily chemical production line online control system according to claim 1 is characterized in that: The fault prediction and remote operation and maintenance module includes the following functions: Based on the equipment operation analysis data, historical operation record trend modeling is performed to obtain equipment operation trend data; Perform historical failure feature pattern matching analysis on key equipment components based on equipment operation trend data to obtain equipment failure prediction data; Perform maintenance strategy optimization calculation based on equipment failure prediction data to obtain equipment predictive maintenance plan data; Based on the equipment predictive maintenance plan data, the intelligent daily chemical production line is remotely operated and maintained to obtain an optimized production control system.

10. The intelligent daily chemical production line online control system according to claim 9 is characterized in that: The matching analysis of historical fault feature patterns of key equipment components based on equipment operation trend data includes: Extract historical fault records from equipment operation trend data, classify fault modes, and obtain equipment fault mode classification data; Perform statistical analysis on characteristic parameters of each failure mode based on equipment failure mode classification data to obtain equipment failure characteristic parameter data; Based on the equipment fault characteristic parameter data, the equipment key components operating status characteristic matching calculation is performed to obtain the equipment fault characteristic matching data; Use equipment failure feature matching data to calculate the probability of failure of key equipment components and obtain equipment failure probability prediction data; Based on the equipment failure probability prediction data, abnormal trend analysis of equipment operation status is performed to obtain equipment abnormal trend analysis data; Use equipment abnormal trend analysis data to evaluate the fault risk level and obtain equipment fault risk level data; The equipment failure probability prediction data and the equipment failure risk level data are integrated to obtain the equipment failure prediction data.

Images