Smart production control method for MES based on cloud computing

Abstract

To provide an intelligent production control method of an MES based on cloud computing.SOLUTION: Acquiring a real-time production data set of an organic chemical product, analyzing the real-time production data set, and identifying a production bottleneck stage, analyzing the real-time production data set based on the production bottleneck stage, and identifying a plurality of optimizable factors, simulating a production process of the organic chemical product based on the plurality of optimizable factors, and determining an adjustment result, adjusting the real-time production data set based on the adjustment result, and determining a bottleneck optimization data set, and controlling a production device based on the bottleneck optimization data set to perform corresponding adjustment.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] This application relates to the technical field of production control, and in particular, to a smart production control method for MES based on cloud computing.

Background Art

[0002] With the development of industrial informatization, by using the MES system (Manufacturing Execution System) to manage and monitor the industrial production process, it helps to improve production efficiency and product quality, and achieve the goals of lean production and information-based smart manufacturing of enterprises.

[0003] When the conventional MES system manages and monitors the production process of organic chemical industrial products, the manufacturing process of organic chemical industrial products is complex, and the products change dynamically frequently during the production process. Therefore, it is difficult for the conventional MES system to flexibly respond to many variables existing during the production process of organic chemical industrial products, resulting in variations in the quality of organic chemical industrial products and even the possibility of causing safety accidents.

Summary of the Invention

Problems to be Solved by the Invention

[0004] To solve the above technical problems, this application provides a smart production control method for MES based on cloud computing.

Means for Solving the Problems

[0005] In a first aspect, this application provides a smart production control method for MES based on cloud computing, acquiring a real-time production dataset of organic chemical industrial products, analyzing the real-time production dataset, and identifying the manufacturing bottleneck stage; analyzing the real-time production dataset based on the manufacturing bottleneck stage and identifying a plurality of optimizable factors; The steps include: simulating the production process of the organic chemical industrial product based on the aforementioned multiple optimizeable factors and determining the adjustment results; Based on the adjustment results, the real-time production dataset is adjusted to determine the bottleneck optimization dataset. The steps include controlling the production equipment and making corresponding adjustments based on the aforementioned bottleneck optimization dataset, The above method, including.

[0006] This application identifies manufacturing bottleneck stages that cause variations in the quality of organic chemical products by analyzing real-time production data sets during the production process of organic chemical products, identifies multiple optimizable factors within the manufacturing bottleneck stages by analyzing the real-time production data sets, determines adjustment results by simulating the production process of organic chemical products based on the multiple optimizable factors, adjusts the real-time production data sets based on the adjustment results to obtain a bottleneck optimization data set, and controls the production equipment based on the bottleneck optimization data set to make corresponding adjustments. This allows for flexible response to many variables present in the production process of organic chemical products, timely and accurate adjustment of production equipment, guarantees the quality of organic chemical products, and reduces the probability of safety accidents.

[0007] Alternatively, the step of analyzing the real-time production dataset and identifying manufacturing bottleneck stages is: The steps include: applying time smoothing to each real-time production data in the real-time production dataset based on the time interval, and determining low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time; The steps include determining the variability index of each real-time production data during the predetermined total analysis time, based on the low-variability smoothed data, according to the real-time production dataset, The steps include determining the relationship fluctuation index between each real-time production data and other real-time production data based on the fluctuation index of each real-time production data, The process includes the step of analyzing the aforementioned relationship fluctuation index for each real-time production data to identify manufacturing bottleneck stages.

[0008] Through the above technical means, time smoothing is applied to each real-time production data based on a predetermined total analysis time and predetermined analysis time to obtain low-variability smoothed data that is easier to analyze and process. Based on the low-variability smoothed data, a relational variation index is determined that can reflect the actual fluctuation status of the current real-time production data under the influence of different real-time production data. Based on the relational variation index, manufacturing bottleneck stages are identified, improving the efficiency of the mathematical analysis process while enabling the relational variation index to more comprehensively reflect the fluctuation status of the real-time production data and improving the accuracy of identifying manufacturing bottleneck stages.

[0009] Alternatively, the step of applying a time smoothing process to each real-time production data in the real-time production dataset based on a predetermined total analysis time and predetermined interval times, and determining low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time, refers to the following formula.

[0010] Number 1 JPEG0007852962000001.jpg46170

[0011] Through the above technical means, a mathematical formula is designed to apply a unified time smoothing process to each real-time production data based on a predetermined total analysis time and predetermined interval time, using mathematical analysis methods. Low-variability smoothed data corresponding to each real-time production data is derived within the predetermined analysis time, reducing the instability of the real-time production data. Subsequently, in the mathematical analysis of the low-variability smoothed data, the sensitivity of the formula to noise is reduced, and the accuracy of the subsequent variation index is improved.

[0012] Alternatively, the step of determining the index of variation for each real-time production data point during the predetermined total analysis time, based on the low-variability smoothed data and according to the real-time production dataset, refers to the following formula:

[0013] Math 2 JPEG0007852962000002.jpg60170

[0014] Through the above technical means, a mathematical formula is designed to quantify the degree of variation of each real-time production data based on low-variability smoothed data using mathematical analysis methods, and a variation index is derived. This allows the variation index to comprehensively and intuitively reflect the changes in real-time production data on the time axis corresponding to the entire predetermined analysis period, thereby improving the accuracy and comprehensiveness of the variation index.

[0015] Alternatively, the step of determining the relationship fluctuation index between each real-time production data and other real-time production data based on the fluctuation index of each real-time production data refers to the following formula:

[0016] Math 3 JPEG0007852962000003.jpg54170

[0017] Through the above technical means, a mathematical formula is designed using mathematical analysis methods based on the fluctuation index of each real-time production data. The influence of other real-time production data on the current real-time production data is comprehensively considered, and the relational fluctuation index is quantified. This allows the relational fluctuation index to more comprehensively reflect the fluctuation status of real-time production data, and subsequently, the accuracy of identifying manufacturing bottleneck stages derived based on the relational fluctuation index is further improved.

[0018] Alternatively, the step of analyzing the aforementioned relationship fluctuation index for each real-time production data to identify manufacturing bottleneck stages includes the following steps: Based on a predetermined relationship variation threshold value, referring to the following formula to identify a set of positions of a plurality of abnormal relationship variation indices in the real-time production dataset

[0019] Number 4 JPEG0007852962000004.jpg41170

[0020] Based on the set of positions, analyzing the real-time production dataset to identify a plurality of abnormal production data Based on the plurality of abnormal production data, identifying the manufacturing bottleneck stage

[0021] By the above technical means, based on the manufacturing bottleneck stage, analyzing the real-time production dataset, identifying a plurality of production data that become bottlenecks related to the manufacturing bottleneck stage, analyzing the plurality of production data that become bottlenecks, clarifying the sensitivity of each production data that becomes a bottleneck to the plurality of abnormal production data, and based on the sensitivity of each production data that becomes a bottleneck to the plurality of abnormal production data and a predetermined sensitivity threshold value, identifying the plurality of optimizable factors

[0022] Alternatively, the step of analyzing the real-time production dataset based on the manufacturing bottleneck stage and identifying a plurality of optimizable factors includes Based on the manufacturing bottleneck stage, analyzing the real-time production dataset to identify a plurality of production data that become bottlenecks related to the manufacturing bottleneck stage Analyzing the plurality of production data that become bottlenecks to clarify the sensitivity of each production data that becomes a bottleneck to the plurality of abnormal production data Based on the sensitivity of each production data that becomes a bottleneck to the plurality of abnormal production data and a predetermined sensitivity threshold value, identifying the plurality of optimizable factors

[0023] Through the above technical means, based on the manufacturing bottleneck stage, multiple bottleneck production data are analyzed to identify multiple bottleneck production data, and by analyzing these multiple bottleneck production data, the sensitivity of each bottleneck production data to multiple abnormal production data is revealed, and multiple optimizable factors are identified based on sensitivity and a predetermined sensitivity threshold. By introducing the concept of sensitivity, multiple bottleneck production data are selected, and from among them, the multiple bottleneck production data with the greatest need for optimization are identified and used as multiple optimizable factors, which helps to improve the efficiency of subsequent production process adjustments and to accurately grasp the results of subsequent production process adjustments.

[0024] Alternatively, the step of analyzing the multiple bottleneck production data and determining the sensitivity of each bottleneck production data to the multiple abnormal production data refers to the following formula.

[0025] Number 5 JPEG0007852962000005.jpg48170

[0026] Through the above technical means, a mathematical formula is designed based on multiple bottleneck production data using mathematical analysis methods, regression analysis is performed on the sensitivity of each bottleneck production data to the multiple abnormal production data, the degree of relationship between the current bottleneck production data and the current abnormal production data is reflected by a predetermined relation index, the error in sensitivity is reduced by a disturbance term, and finally the sensitivity of each bottleneck production data to the multiple abnormal production data is derived, improving the accuracy and comprehensiveness of the degree of relationship.

[0027] Alternatively, the step of simulating the production process of the organic chemical industrial product based on the multiple optimizable factors and determining the adjustment results is: A step of simulating a real-time production process based on predetermined manufacturing process information and the real-time production dataset, The steps include adjusting the real-time production process based on the aforementioned multiple optimizeable factors and simulating the adjusted production process, The process includes the step of analyzing the adjusted production process and determining multiple production data within the adjusted production process as adjustment results.

[0028] Through the above technical means, a real-time production process is simulated based on predetermined manufacturing process information and the real-time production dataset; the real-time production process is adjusted based on multiple optimizeable factors; the adjusted production process is analyzed; and corresponding multiple production data within the adjusted production process are determined as adjustment results. This improves the accuracy of the adjustment results, avoids irreversible consequences from direct adjustments to the actual production process, and provides an intuitive data base for subsequent adjustments to the actual production process.

[0029] Alternatively, the above method is A step of constructing a dynamic dataset by recording data at each time node within the analysis time divided by predetermined time intervals for each production data within the real-time production dataset and the adjustment results, based on the real-time production dataset and the adjustment results, The method further includes the step of identifying, outputting, and making available for viewing by expert technicians the visualized dynamic production data based on the aforementioned dynamic dataset.

[0030] Through the above technical means, real-time production datasets and adjustment results are analyzed, dynamic datasets are identified, and dynamic datasets are visualized using data visualization technology. Visualized dynamic production data is identified, and this visualized dynamic production data is provided for analysis by expert technicians, enabling them to understand the changes in production process data and providing them with intuitive and comprehensive data for improving the manufacturing processes of organic chemical industrial products.

[0031] The following brief description of the accompanying drawings, which are necessary for depicting embodiments or the prior art, is provided below to clearly illustrate the embodiments of this application or the technical means within the prior art. The accompanying drawings described below are merely some embodiments of this application, and a person skilled in the art can obtain other drawings based on these accompanying drawings, assuming no creative activity is required. [Brief explanation of the drawing]

[0032] [Figure 1] This is a schematic diagram illustrating an application scenario provided in one embodiment of this application. [Figure 2] This is a flowchart of a smart production control method for a cloud computing-based MES provided in one embodiment of this application. [Modes for carrying out the invention]

[0033] To further clarify the purpose, technical means, and advantages of this application, the technical means in the embodiments of this application will be described in detail below with reference to the drawings in the embodiments of this application, although it goes without saying that the embodiments described are only some of the embodiments of this application, not all of them. All other embodiments derived from the embodiments of this application, without any creative activity by a person skilled in the art, are all within the scope of protection of this application.

[0034] As used herein, the term "and / or" includes any and all combinations of one or more related components listed together. For example, A and / or B means either A exists alone, A and B exist together, or B exists alone. As used herein, the symbol " / " generally indicates an "or" relationship between the preceding and following related components, unless otherwise specified.

[0035] The embodiments of this application will be described in detail below with reference to the drawings in the specification.

[0036] When managing and monitoring the production process of organic chemical products using conventional MES methods, the manufacturing process of organic chemical products is complex, and the product changes dynamically frequently during the production process. Therefore, conventional MES methods have difficulty flexibly responding to the many variables present in the production process of organic chemical products, which can lead to variations in the quality of organic chemical products and potentially cause safety accidents.

[0037] Based on this, the smart production control method for a cloud computing-based MES provided in this application analyzes a real-time production dataset during the production process of organic chemical products, identifies manufacturing bottleneck stages that cause variations in the quality of organic chemical products, identifies multiple optimizable factors during the manufacturing bottleneck stages by analyzing the real-time production dataset, determines adjustment results by simulating the production process of organic chemical products based on the multiple optimizable factors, adjusts the real-time production dataset based on the adjustment results to obtain a bottleneck optimization dataset, and controls the production equipment based on the bottleneck optimization dataset to make corresponding adjustments. This method flexibly responds to many variables present in the production process of organic chemical products, adjusts the production equipment accurately and in a timely manner, guarantees the quality of organic chemical products, and reduces the probability of safety accidents.

[0038] Figure 1 is a schematic diagram illustrating an application scenario provided in one embodiment of this application. In the production process of organic chemical industrial products, the method provided in this application is used to analyze real-time production data and control production equipment to make corresponding adjustments.

[0039] Specifically, the method provided in this application is installed on any server, which communicates with production equipment, acquires real-time production data sets provided by the production equipment, analyzes the real-time production data sets to identify manufacturing bottleneck stages that cause variations in the quality of organic chemical products, identifies multiple optimizable factors within the manufacturing bottleneck stages by analyzing the real-time production data sets, determines adjustment results by simulating the production process of organic chemical products based on the multiple optimizable factors, adjusts the real-time production data sets based on the adjustment results to obtain a bottleneck optimization data set, and controls the production equipment to make corresponding adjustments based on the bottleneck optimization data set. This allows for flexible response to many variables present in the production process of organic chemical products, timely and accurate adjustment of the production equipment, guarantees the quality of organic chemical products, and reduces the probability of safety accidents.

[0040] For specific implementation methods, please refer to the following embodiments.

[0041] Figure 2 is a flowchart of a smart production control method for a cloud computing-based MES provided in one embodiment of this application. The method of this embodiment can be used for a server in the above scenario. As shown in Figure 2, the method includes the following steps.

[0042] S201: A step to obtain real-time production data for organic chemical industrial products, analyze the real-time production data, and identify bottleneck stages in the manufacturing process.

[0043] Organic chemical industrial products can be chemical products produced from organic compounds through chemical reactions and processing.

[0044] Real-time production data can consist of a series of real-time parameters (temperature, reaction pressure, pH value, etc.) of the current chemical reaction stages of organic chemical products in the production process. Real-time production data can be acquired through various sensors placed on the production line of organic chemical products.

[0045] A manufacturing bottleneck can be a stage in the current production process of organic chemical products that negatively impacts the final quality of the product.

[0046] Specifically, the production process for organic chemical products involves complex chemical reactions, and the various raw materials of the product are extremely sensitive to the reaction conditions during the reaction process. Therefore, if abnormalities occur in the reaction conditions, the quality of the organic chemical products often results in batch failure. Furthermore, because organic chemical products undergo a continuous series of multi-stage chemical reactions during the production process, it is difficult to identify and adjust the stage where the problem occurred in a timely manner, even if a batch of products fails. This leads to variations in the quality of organic chemical products and makes it impossible to maintain a high yield rate.

[0047] By using transit mathematical analysis techniques to analyze real-time production data during the production process of organic chemical industrial products and quantifying the changes in each production data within the same production time series, abnormal bottleneck processes can be identified, providing fundamental data for subsequent production process adjustments.

[0048] S202: A step to analyze real-time production data based on manufacturing bottleneck stages and identify multiple factors that can be optimized.

[0049] The factors that can be optimized may be the corresponding production data that needs to be adjusted in the bottleneck process.

[0050] Specifically, since each manufacturing stage is accompanied by a series of production data, and the cause of identifying a manufacturing bottleneck stage is an anomaly resulting from the interaction of multiple production data, after identifying the manufacturing bottleneck stage, mathematical analysis methods are used to quantify the impact of production data within the real-time production dataset on the manufacturing bottleneck stage, identify multiple corresponding optimizable factors within the manufacturing bottleneck stage, and then provide a basis for adjusting organic chemical industrial products.

[0051] S203: A step in which the production process of an organic chemical industrial product is simulated based on multiple optimizeable factors, and adjustment results are determined.

[0052] The adjustment results may be adjustment values ​​corresponding to multiple optimizeable factors obtained by optimizing multiple optimizeable factors and simulating the production process of organic chemical industrial products.

[0053] Specifically, because organic chemical industrial products are highly sensitive to reaction conditions during the production process, it is difficult to ensure real-time production efficiency by directly adjusting multiple optimizable factors after identifying them. Therefore, based on multiple optimizable factors, the production process of organic chemical industrial products is simulated using process simulation software, such as Aspen Plus (Advanced System for Process Engineering), to determine the values ​​that need to be adjusted for the multiple optimizable factors, and to provide accurate data for subsequent adjustments to the actual production process of organic chemical industrial products.

[0054] S204: A step of adjusting the real-time production dataset based on the adjustment results to determine the bottleneck optimization dataset.

[0055] The bottleneck optimization dataset may be a set containing all production data, both adjusted and unadjusted, after the real-time production dataset has been adjusted based on the adjustment results.

[0056] Specifically, based on adjustment values ​​corresponding to multiple optimizeable factors in the adjustment results, production data within a real-time production dataset is adjusted to obtain a bottleneck optimization dataset containing all adjusted and unadjusted production data, providing a data foundation for accurate adjustments in subsequent control of physical production equipment.

[0057] S205: A step to control production equipment and make corresponding adjustments based on a bottleneck optimization dataset.

[0058] Production equipment can be a series of physical pieces of equipment on a production line for producing organic chemical industrial products.

[0059] Specifically, after determining the bottleneck optimization dataset, the system communicates with physical equipment via a cloud computing platform. Based on the bottleneck optimization dataset, the production equipment is remotely controlled, and the equipment makes corresponding adjustments in real time to ensure product acceptance rates.

[0060] The method provided in this embodiment analyzes real-time production data during the production process of organic chemical products to identify manufacturing bottleneck stages that cause variations in the quality of organic chemical products, identifies multiple optimizable factors within the manufacturing bottleneck stages by analyzing the real-time production data, determines adjustment results by simulating the production process of organic chemical products based on the multiple optimizable factors, adjusts the real-time production data based on the adjustment results to obtain a bottleneck optimization data set, and controls the production equipment based on the bottleneck optimization data set to make corresponding adjustments. This allows for flexible response to many variables present in the production process of organic chemical products, timely and accurate adjustment of production equipment, guarantees the quality of organic chemical products, and reduces the probability of safety accidents.

[0061] In some embodiments, time smoothing is applied to each real-time production data in the real-time production dataset based on a predetermined total analysis time and predetermined interval times; low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time is determined; based on the low-variability smoothed data, the variability index of each real-time production data at a predetermined total analysis time is determined according to the real-time production dataset; based on the variability index of each real-time production data, the relationship variability index between each real-time production data and other real-time production data is determined; and the relationship variability index of each real-time production data is analyzed to identify manufacturing bottleneck stages.

[0062] The predetermined total analysis time may be a pre-set analysis time for real-time production data. The predetermined total analysis time can be set according to the actual production situation.

[0063] The predetermined time intervals may be used to identify time information for different time nodes by dividing the pre-set total analysis time.

[0064] Low-variability, smoothed data can be a data source that clearly reflects the trend of change in real-time production data.

[0065] The volatility index can reflect the intensity of fluctuations in real-time production data.

[0066] The relationship fluctuation index can be data that shows the extent to which current real-time production data fluctuations are influenced by fluctuations in other real-time production data.

[0067] Specifically, directly acquired production datasets contain numerous different types of production data, each influenced by different factors, and the criteria for measuring fluctuations vary significantly, making it difficult to perform a unified mathematical analysis. By dividing a predetermined analysis time into multiple time nodes with predetermined time intervals, and applying time smoothing to the production data corresponding to each time node in the real-time production dataset, low-fluctuation smoothed data is obtained. This removes noise from the production data while maintaining the fluctuation characteristics of the production data, reducing sensitivity to noise in the subsequent mathematical analysis process. Furthermore, the low-fluctuation smoothed data not only reflects the changing trends of real-time production data, but is also easier to analyze and process, contributing to improved efficiency in subsequent mathematical analysis.

[0068] After obtaining low-variability smoothed data, a quantitatively obtained variability index using mathematical analysis methods is used to reflect the current variability of real-time variability data on a time axis corresponding to a predetermined analysis time. Furthermore, since different production data are not independent of each other, different production data influence each other in the production process of organic chemical products. The variability of real-time production data must be considered based on the variability index, taking into account the influence of other real-time production data. Using mathematical analysis methods, a quantitatively obtained relational variability index is used based on the variability index to reflect the actual variability of current real-time production data under the influence of different real-time production data. By analyzing abnormal variability within the relational variability index, the corresponding manufacturing bottleneck stage is identified.

[0069] The method provided in this embodiment applies a time smoothing process to each real-time production data based on a predetermined total analysis time and a predetermined analysis time to obtain low-variability smoothed data that is easier to analyze and process. Based on the low-variability smoothed data, a relational variation index is determined that can reflect the actual fluctuations of the current real-time production data under the influence of different real-time production data. Based on the relational variation index, manufacturing bottleneck stages are identified, improving the efficiency of the mathematical analysis process while enabling the relational variation index to more comprehensively reflect the fluctuations of the real-time production data and improving the accuracy of identifying manufacturing bottleneck stages.

[0070] In some embodiments, applying a time smoothing process to each real-time production data in a real-time production dataset based on a predetermined total analysis time and a predetermined interval time, and determining low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time, is described by the following equation (1).

[0071] Number 6 JPEG0007852962000006.jpg46170

[0072] Specifically, equation (1) applies a uniform smoothing process to the real-time production data values ​​on the most recent N time nodes to obtain a low-variability smoothed value for each real-time production data over a predetermined total analysis time, thereby reducing the instability of the real-time production data.

[0073] The method provided in this embodiment uses mathematical analysis techniques to design a formula for applying a unified time smoothing process to each real-time production data based on a predetermined total analysis time and predetermined interval times. Within the predetermined analysis time, low-variability smoothed data corresponding to each real-time production data is derived, reducing the instability of the real-time production data. Subsequently, in the mathematical analysis of the low-variability smoothed data, the sensitivity of the formula to noise is reduced, and the accuracy of the subsequent variation index is improved.

[0074] In some embodiments, determining the index of variation for each real-time production data point over a predetermined total analysis time, according to the real-time production dataset, based on the low-variability smoothed data, is shown in equation (2) below.

[0075] Number 7 JPEG0007852962000007.jpg60170

[0076] Specifically, the difference between the specific data for each real-time production data and the corresponding low-variability smoothed data is quantified using equation (2) Xi,j-LDi,t. The degree of variation for each real-time production data is reflected by the standard difference between the specific data for each real-time production data and the corresponding low-variability smoothed data, and the variation index is calculated. As a result, the variation index can comprehensively and intuitively reflect the changes in real-time production data over a time axis corresponding to the entire predetermined analysis period.

[0077] The method provided in this embodiment uses mathematical analysis techniques to design a formula for quantifying the degree of variation of each real-time production data based on low-variability smoothed data, and derives a variation index. This allows the variation index to comprehensively and intuitively reflect the changes in real-time production data on a time axis corresponding to the entire predetermined analysis period, thereby improving the accuracy and comprehensiveness of the variation index.

[0078] In some embodiments, determining the relationship fluctuation index between each real-time production data and other real-time production data based on the fluctuation index of each real-time production data refers to the following equation (3).

[0079] Number 8 JPEG0007852962000008.jpg68170

[0080] A given relational data matrix can be a pre-configured symmetric matrix for storing the correlation coefficient between any two real-time production data sets. The diagonal elements of the given relational data matrix are zero. The given relational data matrix can be obtained by analyzing historical data.

[0081] JPEG0007852962000009.jpg45170

[0082] The method provided in this embodiment allows for the design of a mathematical formula using mathematical analysis techniques based on the variability index of each real-time production data, comprehensively considering the impact of other real-time production data on the current real-time production data, quantifying the relational variability index, enabling the relational variability index to more comprehensively reflect the variability of real-time production data, and further improving the accuracy of identifying manufacturing bottleneck stages derived based on the relational variability index.

[0083] In some embodiments, based on a predetermined relational variation threshold, a set of locations of multiple anomalous relational variation indices within a real-time production dataset is identified by referring to the following equation (4).

[0084] Number 9 JPEG0007852962000010.jpg40170

[0085] A predetermined relational fluctuation threshold can be a set threshold data used to determine whether or not there is an abnormality in the relational fluctuation index. A predetermined relational fluctuation threshold can be obtained by analyzing historical data.

[0086] An abnormal relational fluctuation index may be an abnormal data point among several current relational fluctuation indices that is greater than a predetermined relational fluctuation threshold.

[0087] A position set can be a collection of position subscripts for real-time production data corresponding to multiple anomalous relational variation indices within a real-time production dataset.

[0088] JPEG0007852962000011.jpg72170

[0089] The method provided in this embodiment uses mathematical analysis techniques to design a formula that determines whether or not there is an anomaly in a predetermined anomaly in relational fluctuation index, assigns values ​​to indicator variables and selects them to extract position subscripts of real-time production data corresponding to the anomaly in relational fluctuation index, integrates multiple position subscripts to obtain a position set, identifies manufacturing bottleneck stages through multiple anomaly production data corresponding to the position set, and makes the process of identifying manufacturing bottleneck stages faster and more accurate.

[0090] In some embodiments, a real-time production dataset is analyzed based on a manufacturing bottleneck stage to identify multiple bottleneck production data associated with the manufacturing bottleneck stage, the multiple bottleneck production data are analyzed to determine the sensitivity of each bottleneck production data to multiple abnormal production data, and the multiple optimizable factors are identified based on the sensitivity of each bottleneck production data to the multiple abnormal production data and a predetermined sensitivity threshold.

[0091] The production data that acts as a bottleneck could be multiple real-time production data points that have a high correlation with the production bottleneck.

[0092] Sensitivity can be used to indicate how much current bottleneck production data is affected by multiple anomaly production data.

[0093] A predetermined sensitivity threshold can be used to determine whether adjusting the current bottleneck production data has a significant effect on multiple anomalous production data. This predetermined sensitivity threshold can be obtained by analyzing historical data.

[0094] Specifically, in the manufacturing process of organic chemical products, a series of highly correlated real-time production data exist at each stage, and these real-time production data influence each other. However, these real-time production data cannot be directly adjusted, resulting in low adjustment efficiency on the one hand, and difficulty in stably controlling the adjustment results by directly adjusting the above real-time production data in batches on the other hand. Therefore, based on the manufacturing process of organic chemical products, according to the manufacturing bottleneck stage, a real-time production dataset is analyzed to identify multiple bottleneck production data corresponding to the manufacturing bottleneck stage. Using mathematical analysis methods, multiple bottleneck production data are analyzed to clarify the sensitivity of each bottleneck production data to the aforementioned multiple abnormal production data. If the sensitivity of multiple bottleneck production data is greater than a predetermined sensitivity threshold, it is shown that adjusting these bottleneck production data can effectively affect the multiple abnormal production data, and these bottleneck production data are determined as multiple optimizable factors.

[0095] The method provided in this embodiment analyzes multiple bottleneck production data based on the manufacturing bottleneck stage, identifies multiple bottleneck production data, and by analyzing these multiple bottleneck production data, reveals the sensitivity of each bottleneck production data to multiple abnormal production data, and identifies multiple optimizable factors based on sensitivity and a predetermined sensitivity threshold. By introducing the concept of sensitivity, multiple bottleneck production data can be selected, and from among them, multiple bottleneck production data with the highest need for optimization can be identified and used as multiple optimizable factors, thereby improving the efficiency of subsequent production process adjustments and helping to accurately grasp the results of subsequent production process adjustments.

[0096] In some embodiments, analyzing multiple bottleneck production data and determining the sensitivity of each bottleneck production data to multiple abnormal production data can be done by referring to equation (5):

[0097] Number 10 JPEG0007852962000012.jpg47170

[0098] The predetermined intercept term may be a constant term representing the sensitivity threshold. The predetermined intercept term can be obtained by analyzing historical data.

[0099] The predetermined disturbance term may be a constant term representing the average calculation error of the sensitivity. The predetermined disturbance term can be obtained by analyzing historical data.

[0100] A predetermined correlation index may represent the degree to which bottleneck production data is affected by the corresponding abnormal production data. This predetermined correlation index can be obtained by analyzing historical data.

[0101] Specifically, based on multiple bottleneck production data, a regression analysis is performed on sensitivity using equation (5), and the reference value of the result of equation (5) is limited by a predetermined intercept term to prevent the sensitivity from deviating from the normal range. The degree of relationship between the current bottleneck production data and the current abnormal production data is reflected through a predetermined relation index, and the error in sensitivity is reduced by a disturbance term, ultimately obtaining the sensitivity of each bottleneck production data to multiple abnormal production data.

[0102] The method provided in this embodiment involves designing a mathematical formula based on multiple bottleneck production data using mathematical analysis techniques, performing regression analysis on the sensitivity of each bottleneck production data to the multiple abnormal production data, reflecting the degree of relationship between the current bottleneck production data and the current abnormal production data using a predetermined relation index, reducing the error in sensitivity using a disturbance term, and finally deriving the sensitivity of each bottleneck production data to the multiple abnormal production data, thereby improving the accuracy and comprehensiveness of the degree of relationship.

[0103] In some embodiments, a real-time production process is simulated based on predetermined manufacturing process information and a real-time production dataset; the real-time production process is adjusted based on multiple optimizeable factors; the adjusted production process is simulated; the adjusted production process is analyzed; and multiple production data within the adjusted production process are determined as adjustment results.

[0104] The specified manufacturing process information may be the manufacturing process for current organic chemical products. The specified manufacturing process information may be provided by specialist staff.

[0105] A real-time production process can be a virtual real-time production process simulated based on a real-time production dataset.

[0106] The adjusted production process may be a virtual production process obtained after adjusting the real-time production process based on multiple optimizeable factors.

[0107] Specifically, process simulation software, such as Aspen Plus, is used to simulate a real-time production process based on predetermined manufacturing process information and real-time production datasets. A virtual process of the real-time production process is obtained, and based on multiple optimizable factors, the virtual process of the real-time production process is adjusted using the process simulation software to simulate the adjusted production process. A series of production data from the adjusted production process is extracted, and the above series of production data is finalized as the adjustment result, providing an accurate data base for subsequent adjustments to the actual production process.

[0108] The method provided in this embodiment simulates a real-time production process based on predetermined manufacturing process information and the real-time production dataset, adjusts the real-time production process based on multiple optimizeable factors, analyzes the adjusted production process, and confirms the corresponding multiple production data within the adjusted production process as adjustment results, thereby improving the accuracy of the adjustment results, avoiding irreversible consequences from direct adjustments to the actual production process, and providing an intuitive data base for subsequent adjustments to the actual production process.

[0109] In some embodiments, based on a real-time production dataset and adjustment results, data is recorded for each production data within the real-time production data and adjustment results at each time node within an analysis time divided by predetermined time intervals, a dynamic dataset is constructed, and based on the dynamic dataset, visualized dynamic production data is identified, output, and made available for viewing by expert technicians.

[0110] A dynamic dataset can be a dataset in which real-time production data and production data within adjustment results change along with the production time series.

[0111] Visualized dynamic production data can be data obtained after applying visualization processing to dynamic production data, making it possible for stakeholders to directly and interactively analyze it.

[0112] A specialist technician could be responsible for coordinating the manufacturing process of organic chemical products.

[0113] Specifically, a data visualization library, such as D3.js (Data-Driven Documents), is used to perform dynamic visualization on dynamic datasets, obtaining visualized dynamic production data. This visualized dynamic production data is then output through a human-computer interaction device, such as a high-resolution touchscreen, enabling expert technicians to perform interactive analysis.

[0114] The method provided in this embodiment analyzes real-time production datasets and adjustment results, identifies dynamic datasets, visualizes the dynamic datasets using data visualization technology, identifies visualized dynamic production data, provides the visualized dynamic production data for analysis by expert technicians, enables experts to understand the changes in production process data, and provides expert technicians with intuitive and comprehensive data for improving the manufacturing process of organic chemical industrial products.

Claims

1. A smart production control method for MES based on cloud computing, The steps include obtaining a real-time production dataset of organic chemical industrial products, analyzing the real-time production dataset, and identifying bottleneck stages in the manufacturing process. The steps include analyzing the real-time production dataset based on the aforementioned manufacturing bottleneck stage and identifying multiple factors that can be optimized, The steps include: simulating the production process of the organic chemical industrial product based on the aforementioned multiple optimizeable factors and determining the adjustment results; Based on the adjustment results, the real-time production dataset is adjusted to determine the bottleneck optimization dataset. The steps include controlling the production equipment and making corresponding adjustments based on the aforementioned bottleneck optimization dataset, The step of analyzing the real-time production dataset and identifying manufacturing bottleneck stages is: The steps include: applying a time smoothing process to each real-time production data in the real-time production dataset based on a predetermined total analysis time and predetermined interval times, and determining low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time; The steps include determining the variability index of each real-time production data during the predetermined total analysis time, based on the low-variability smoothed data, according to the real-time production dataset, The steps include determining the relationship fluctuation index between each of the aforementioned real-time production data and other real-time production data based on the fluctuation index of each of the aforementioned real-time production data, The step includes analyzing the relationship fluctuation index of each of the real-time production data to identify the manufacturing bottleneck stage, The step of applying time smoothing to each real-time production data in the real-time production dataset based on a predetermined total analysis time and predetermined interval times, and determining low-variability smoothed data corresponding to each real-time production data at a predetermined analysis time, refers to the following formula: Number 1 The step of determining the index of variation for each real-time production data during the predetermined total analysis time, based on the low-variability smoothed data and according to the real-time production dataset, is performed by referring to the following formula: Math 2 The step of determining the relationship fluctuation index between each real-time production data and other real-time production data based on the fluctuation index of each real-time production data refers to the following formula: Math 3 The step of analyzing the aforementioned relationship fluctuation index for each real-time production data to identify manufacturing bottleneck stages is: The steps include identifying a set of locations of multiple abnormal relational variation indices within the real-time production dataset by referring to the following formula, based on a predetermined relational variation threshold, Math 4 The steps include: analyzing the real-time production dataset based on the aforementioned location set to identify multiple abnormal production data; The step includes identifying the manufacturing bottleneck stage based on the aforementioned multiple abnormal production data, The step of analyzing the real-time production dataset based on the manufacturing bottleneck stage and identifying multiple optimizable factors is: Based on the aforementioned manufacturing bottleneck stage, the step of analyzing a real-time production dataset and identifying multiple bottleneck production data related to the aforementioned manufacturing bottleneck stage, The steps include: analyzing the multiple bottleneck production data and clarifying the sensitivity of each bottleneck production data to the multiple abnormal production data; The step includes identifying a plurality of optimizable factors based on the sensitivity of each bottleneck production data to the plurality of abnormal production data and a predetermined sensitivity threshold, The step of analyzing the multiple bottleneck production data and clarifying the sensitivity of each bottleneck production data to the multiple abnormal production data refers to the following formula: Number 5 A method characterized by the following features.

2. The step of simulating the production process of the organic chemical industrial product based on the aforementioned multiple optimizeable factors and determining the adjustment results is: A step of simulating a real-time production process based on predetermined manufacturing process information and the real-time production dataset, The steps include adjusting the real-time production process based on the aforementioned multiple optimizeable factors and simulating the adjusted production process, The steps include analyzing the adjusted production process and determining multiple production data within the adjusted process as adjustment results, The method according to claim 1, characterized by including

3. Based on the real-time production dataset and the adjustment results, the step of recording the data at each time node within the analysis time divided by the predetermined time intervals for each production data in the real-time production data and the adjustment results, thereby constructing a dynamic dataset; Based on the aforementioned dynamic dataset, the steps include identifying and outputting visualized dynamic production data and making it available for viewing by expert technicians, The method according to claim 2, further comprising:

Images