Sodium dihydrogen phosphate intelligent manufacturing monitoring system based on Internet of Things

Abstract

The invention relates to the technical field of intelligent manufacturing monitoring, in particular to a sodium dihydrogen phosphate intelligent manufacturing monitoring system based on the Internet of Things, which comprises a parameter analysis module, a data hierarchical scheduling module, an abnormal linkage identification module, a trend synchronous mapping module and a storage partition optimization module. According to the method, the spatial distribution characteristics and the fluctuation frequency of the manufacturing parameters are calculated, the multi-level behavior model is constructed, the monitoring point data value recognition capability is improved, parameter region differentiation division is achieved, the data storage frequency is adjusted according to the fluctuation level, distribution according to needs is achieved, redundancy is reduced, the manufacturing characteristic parameters are extracted, and monitoring point logic linkage is analyzed; the abnormal attribution efficiency is improved, a data linkage graph is constructed based on trend consistency, production stability changes are perceived, a storage structure is optimized in combination with spatial layout, the data safety and response efficiency of a key area are enhanced, monitoring is converted from global homogeneous collection to key regulation and control, and the overall efficiency and system toughness are improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent manufacturing monitoring technology, and in particular to an intelligent manufacturing monitoring system for sodium dihydrogen phosphate based on the Internet of Things. Background Technology

[0002] The field of intelligent manufacturing monitoring technology encompasses the comprehensive monitoring, analysis, and management of the manufacturing process through information technology and intelligent technologies. Its core content involves the collection, processing, and analysis of various data generated during the manufacturing process to improve production efficiency, reduce production costs, and ensure product quality. Within this technical field, research focuses on how to utilize modern information technologies such as the Internet of Things, big data, and cloud computing to enhance the automation and intelligence levels of production lines, thereby achieving functions such as real-time monitoring, fault early warning, and predictive maintenance. Through the comprehensive application of these technologies, intelligent manufacturing monitoring can improve overall production efficiency and reliability in various industrial scenarios.

[0003] The sodium dihydrogen phosphate intelligent manufacturing monitoring system is specifically designed for intelligent monitoring and optimization in the sodium dihydrogen phosphate production process. It primarily addresses the real-time monitoring and automatic adjustment of key parameters such as temperature, pressure, and flow rate during production. This is achieved by utilizing sensors and intelligent data acquisition to collect various data points in real time during production. Combined with data analysis and processing technologies, the system automatically alarms and adjusts for any abnormalities that occur during production. Furthermore, through real-time monitoring and intelligent decision-making, the system enables refined management and optimization of the production process, thereby improving the production efficiency and stability of sodium dihydrogen phosphate.

[0004] While existing technologies possess the capability for real-time monitoring of core parameters such as temperature, pressure, and flow rate, they lack refined management in parameter hierarchy classification and data storage. The common practice of using the same-frequency storage strategy for all monitoring points results in inflexible allocation of data processing resources, leading to a large amount of redundant data in areas with low volatility, increasing system load and impacting processing efficiency. Regarding anomaly identification in the manufacturing process, current technologies primarily focus on absolute value warnings for single-point data, lacking in-depth analysis of the behavioral linkages between multiple monitoring points within similar manufacturing units. This easily overlooks potential systemic anomaly trends, resulting in delayed warning responses. In terms of synchronous analysis of data trends, existing systems lack mechanisms to judge the consistency of fluctuation trends among monitoring point data, making it difficult to construct multi-point collaborative evolution models and limiting the ability to proactively control changes in manufacturing status. Furthermore, there is still a lack of coordination between the spatial distribution of monitoring points and the data storage architecture. The storage of data from each monitoring point does not take into account its location and role in the manufacturing process, resulting in a situation where regional information redundancy and weak storage in key areas coexist. Long-term operation has led to monitoring blind spots and data security risks, resulting in problems such as high monitoring load, deviation of anomaly diagnosis from reality, and data scheduling imbalance, which restricts the further improvement of monitoring efficiency and system stability. Summary of the Invention

[0005] To address the technical problems existing in the prior art, this invention provides an IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate. The technical solution is as follows:

[0006] On the one hand, an IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate was provided, which includes:

[0007] Based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, the parameter analysis module calculates the fluctuation range and frequency of manufacturing parameters, divides the parameters into three categories: high, medium, and low fluctuation, analyzes the distribution density of monitoring points, filters dense areas, and obtains the parameter distribution analysis results.

[0008] Based on the parameter distribution analysis results, the data hierarchical scheduling module sets the data storage priority, adjusts the data storage frequency of differentiated fluctuation areas, sets high fluctuation areas as high frequency storage, medium fluctuation areas as medium frequency storage, and low fluctuation areas as low frequency storage, and obtains a data storage frequency partition dataset.

[0009] The abnormal linkage identification module calls the data storage frequency partition dataset, extracts manufacturing feature parameters, filters monitoring points in the same type of manufacturing unit based on feature differences, and obtains abnormal linkage comparison relationships.

[0010] Based on the aforementioned abnormal linkage comparison relationship, the trend synchronization mapping module analyzes the consistency of the fluctuation trend of manufacturing parameters, judges the correlation of monitoring point data, sets a correlation threshold, filters related monitoring points, and obtains the trend of production stability changes.

[0011] As a further aspect of the present invention, the parameter distribution analysis results include high fluctuation regions, medium fluctuation regions, and low fluctuation regions; the data storage frequency partition dataset includes high frequency storage regions, medium frequency storage regions, and low frequency storage regions; the abnormal linkage comparison relationship includes monitoring point units with similar temperature, monitoring point units with similar pressure, and monitoring point units with similar flow rate; and the production stability change trend includes manufacturing parameter fluctuation trend, monitoring point consistency trend, correlation threshold, and correlated monitoring points.

[0012] As a further aspect of the present invention, the parameter distribution analysis module includes:

[0013] The fluctuation range calculation submodule measures and records manufacturing parameter values ​​based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, analyzes the fluctuation range and frequency of change between each monitoring point, and obtains parameter fluctuation values ​​based on the fluctuation range and frequency of change.

[0014] The fluctuation classification submodule calls the parameter fluctuation value, identifies the fluctuation category based on the fluctuation characteristics of adjacent monitoring points, determines the fluctuation range to which the monitoring point belongs, sets the fluctuation classification standard, divides the monitoring points into three categories of areas: high fluctuation, medium fluctuation, and low fluctuation, analyzes the distribution of monitoring points in the area, and generates a monitoring point fluctuation distribution overview.

[0015] The density filtering submodule calls the monitoring point fluctuation distribution overview, calculates the spatial density of monitoring points in the fluctuation area, filters dense areas, and obtains parameter distribution analysis results.

[0016] As a further aspect of the present invention, the spatial density of the monitoring points in the fluctuation region is determined by the formula:

[0017]

[0018] Where ρ represents the spatial density of monitoring points in the fluctuation region, z i Let μ represent the parameter value of the i-th monitoring point, μ represent the mean of the parameter, σ represent the standard deviation of the parameter, p represent the weighting coefficient, A represent the area of ​​the monitoring region, and n represent the total number of monitoring points.

[0019] As a further aspect of the present invention, the data hierarchical scheduling module includes:

[0020] The storage priority setting submodule calls the parameter distribution analysis results, sets the storage priority according to the parameter fluctuation area, sorts them according to the degree of fluctuation, and obtains the monitoring point storage priority data;

[0021] The storage frequency adjustment submodule calls the storage priority data of the monitoring points, adjusts the storage frequency according to the priority, sets high frequency storage for high fluctuation areas, medium frequency storage for medium fluctuation areas, and low frequency storage for low fluctuation areas, analyzes the changes in storage frequency, adjusts the size of the storage dataset, and obtains the data storage frequency partition dataset.

[0022] As a further aspect of the present invention, the abnormal linkage identification module includes:

[0023] The manufacturing feature recognition submodule extracts the temperature, pressure, and flow rate of the monitoring points based on the data storage frequency partition dataset to obtain a set of manufacturing feature parameters.

[0024] The monitoring point classification submodule calls the manufacturing feature parameter set, analyzes the differences in manufacturing features based on temperature, pressure, and flow rate, filters monitoring points with similar feature values, determines the category to which the monitoring points belong, divides the same manufacturing feature units, optimizes the classification boundary, and obtains the distribution of manufacturing feature units.

[0025] The linkage comparison generation submodule calls the manufacturing feature unit distribution, filters monitoring points within the same unit, and obtains abnormal linkage comparison relationships.

[0026] As a further aspect of the present invention, the trend synchronization mapping module includes:

[0027] The monitoring point trend analysis submodule calls the abnormal linkage comparison relationship to analyze the fluctuation trend of manufacturing parameters, extract time series data, and obtain the trend change characteristics of monitoring points;

[0028] The data correlation judgment submodule calls the trend change characteristics of the monitoring points, analyzes the consistency of the fluctuation trend of manufacturing parameters, calculates the data correlation of the monitoring points, sets the correlation threshold, filters the data correlation of the monitoring points, removes low correlation monitoring points, optimizes the data matching relationship, and obtains the monitoring point correlation filtering results.

[0029] The stability trend filtering submodule calls the monitoring point association filtering results, filters the trend change data of the associated monitoring points, identifies the magnitude of production trend changes, and obtains the production stability change trend.

[0030] As a further aspect of the present invention, the magnitude of the change in production trend is determined by the following formula:

[0031]

[0032] Where ΔT represents the magnitude of change in production trend, T j T represents the production trend value at time j. j-1 α represents the production trend value at time j-1, α represents the weighting coefficient of the trend change magnitude, and m represents the total number of time periods.

[0033] As a further aspect of the present invention, the system also includes a storage partition optimization module:

[0034] Based on the trend of production stability changes, the storage partition optimization module extracts the spatial layout information of monitoring points, identifies the corresponding positions between monitoring points, divides storage partitions according to layout characteristics, adjusts the storage partition structure, and obtains a secure storage solution for intelligent manufacturing monitoring data.

[0035] The intelligent manufacturing monitoring data security storage solution includes storage partitioning, storage partition structure adjustment, extraction of monitoring point spatial layout information, and identification of corresponding monitoring point locations.

[0036] As a further aspect of the present invention, the storage partition optimization module includes:

[0037] The spatial layout extraction submodule calls the production stability change trend, extracts the spatial layout data of the monitoring points, identifies the corresponding location of the monitoring points, and obtains the spatial layout information of the monitoring points.

[0038] The storage partitioning submodule calls the spatial layout information of the monitoring points, identifies layout features based on the corresponding locations of the monitoring points, filters dense areas of monitoring points, divides storage partitions, sets differentiated storage partition boundaries, adjusts the distribution balance of monitoring points, and obtains the storage partitioning results.

[0039] The partition structure adjustment submodule calls the storage partitioning results, adjusts the storage partition structure, optimizes the data mapping of monitoring points, and obtains a secure storage solution for intelligent manufacturing monitoring data.

[0040] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following:

[0041] By calculating the spatial distribution characteristics of sodium dihydrogen phosphate manufacturing parameters, the fluctuation amplitude and frequency of each parameter are clarified, enabling the construction of a multi-level parameter behavior model during the manufacturing process. This enhances the ability to identify the value of data from different monitoring points. Monitoring areas are selected based on the density of fluctuation characteristics, completing the differentiated division of parameter regions and providing a refined basis for subsequent processing. The data storage frequency is dynamically adjusted according to the parameter fluctuation level, and different data recording strategies are formulated based on high, medium, and low fluctuation levels to achieve on-demand allocation of data resources, avoid redundant storage burdens, and improve storage efficiency. By extracting and analyzing the differences in manufacturing characteristic parameters, logical linkage relationships between monitoring points are established within the manufacturing unit, enabling rapid attribution and localization of abnormal behaviors. Furthermore, through trend consistency analysis, data linkage maps are established for monitoring points with similar fluctuation trends, forming a dynamic perception capability of the evolution trend of production stability and assisting in the forward-looking judgment of manufacturing status. By comprehensively considering the spatial layout information of monitoring points, corresponding areas are reconstructed by partitioning, and the spatial distribution of monitoring data is optimized in accordance with trend sensitivity, enabling high-security and high-real-time processing of monitoring information in key areas and improving the resilience and adaptability of the data storage structure. The overall processing logic is based on parameter dynamic behavior recognition, zoning strategy formulation, linkage trend analysis and spatial layout optimization, which enables the monitoring of the manufacturing process to shift from homogeneous collection across the entire domain to intelligent control focusing on key dimensions, achieving multi-dimensional collaborative optimization in terms of accuracy, efficiency and resource utilization. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate provided in an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of the system framework of the present invention;

[0045] Figure 3 This is a flowchart of the parameter analysis module in this invention;

[0046] Figure 4 This is a flowchart of the data hierarchical scheduling module in this invention;

[0047] Figure 5 This is a flowchart of the abnormal linkage identification module in this invention;

[0048] Figure 6This is a flowchart of the trend synchronization mapping module in this invention;

[0049] Figure 7 This is a flowchart of the storage partition optimization module in this invention. Detailed Implementation

[0050] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0051] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0052] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, their intended meanings are consistent.

[0053] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0054] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0055] This invention provides an IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate, such as... Figure 1-2 The diagram shown illustrates an IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate. The system includes:

[0056] Based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, the parameter analysis module calculates the fluctuation range and frequency of manufacturing parameters, divides the parameters into three categories: high, medium, and low fluctuation, analyzes the distribution density of monitoring points, filters dense areas, and obtains the parameter distribution analysis results.

[0057] Based on the parameter distribution analysis results, the data hierarchical scheduling module sets the data storage priority, adjusts the data storage frequency of differentiated fluctuation areas, sets high fluctuation areas as high frequency storage, medium fluctuation areas as medium frequency storage, and low fluctuation areas as low frequency storage, and obtains the data storage frequency partition dataset.

[0058] The abnormal linkage identification module calls the data storage frequency partition dataset, extracts manufacturing feature parameters, filters monitoring points within the same type of manufacturing unit based on feature differences, and obtains abnormal linkage comparison relationships.

[0059] The trend synchronization mapping module analyzes the consistency of manufacturing parameter fluctuation trends based on the abnormal linkage comparison relationship, judges the correlation of monitoring point data, sets correlation thresholds, filters related monitoring points, and obtains the trend of production stability changes.

[0060] Based on the trend of production stability changes, the storage partition optimization module extracts the spatial layout information of monitoring points, identifies the corresponding positions between monitoring points, divides storage partitions according to layout characteristics, adjusts the storage partition structure, and obtains a secure storage solution for intelligent manufacturing monitoring data.

[0061] The parameter distribution analysis results include high fluctuation areas, medium fluctuation areas, and low fluctuation areas. The data storage frequency partition dataset includes high frequency storage areas, medium frequency storage areas, and low frequency storage areas. The abnormal linkage comparison relationship includes monitoring point units with similar temperature, similar pressure, and similar flow. The production stability change trend includes manufacturing parameter fluctuation trend, monitoring point consistency trend, correlation threshold, and related monitoring points. The intelligent manufacturing monitoring data security storage scheme includes storage partitioning, storage partition structure adjustment, monitoring point spatial layout information extraction, and monitoring point corresponding location identification.

[0062] Specifically, such as Figure 2 , 3 As shown, the parameter distribution analysis module includes:

[0063] The fluctuation range calculation submodule measures and records manufacturing parameter values ​​based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, analyzes the fluctuation range and frequency of change between each monitoring point, and obtains parameter fluctuation values ​​based on the fluctuation range and frequency of change.

[0064] Based on the spatial distribution characteristics of the monitoring points—comprising temperature, pressure, and flow sensors—deployed at key nodes of the sodium dihydrogen phosphate production line, statistical analysis was performed on manufacturing parameter values ​​collected and recorded at a frequency of once per minute over the past 24 hours. For each monitoring point, the difference between its maximum and minimum values ​​within the monitoring period was calculated as the fluctuation range. The frequency of change was assessed by recording the number of times the parameter value exceeded the historical mean ± 3 standard deviations. For example, if the temperature at monitoring point A reached a maximum of 35°C and a minimum of 30°C within 24 hours, then... The temperature fluctuation range is 5℃. If the number of times the temperature value exceeds the historical average ± 3 standard deviations is 10, then the change frequency is 10 times / 24 hours. By weighted averaging the data of the two dimensions of fluctuation range and change frequency, the weighting coefficients are 0.6 and 0.4, respectively. For example, the fluctuation range of monitoring point A is normalized to 0.5, and the change frequency is normalized to 0.7. Then its parameter fluctuation value is 0.5 + 0.6 + 0.7 + 0.4 = 0.58. Thus, the parameter fluctuation values ​​of all monitoring points are obtained for subsequent fluctuation characteristic analysis.

[0065] The fluctuation classification submodule calls the fluctuation value parameter, identifies the fluctuation category based on the fluctuation characteristics of adjacent monitoring points, determines the fluctuation range to which the monitoring point belongs, sets the fluctuation classification standard, divides the monitoring points into three categories of areas: high fluctuation, medium fluctuation, and low fluctuation, analyzes the distribution of monitoring points in the area, and generates a monitoring point fluctuation distribution overview.

[0066] The system receives parameter fluctuation values ​​from each monitoring point and examines the differences in parameter fluctuation values ​​between adjacent monitoring points. If the difference in fluctuation values ​​between two adjacent monitoring points is less than 0.1, their fluctuation characteristics are considered similar. Based on a pre-set fluctuation classification standard, the monitoring points are divided into three fluctuation ranges: parameter fluctuation values ​​greater than 0.7 are classified as high fluctuation ranges, parameter fluctuation values ​​between 0.4 and 0.7 are classified as medium fluctuation ranges, and parameter fluctuation values ​​less than 0.4 are classified as low fluctuation ranges. For example, if the fluctuation value of monitoring point B is 0.85, it is classified as a high fluctuation range, and the fluctuation value of monitoring point C is... If the fluctuation value of monitoring point D is 0.55, it is classified as a medium fluctuation region. If the fluctuation value of monitoring point D is 0.2, it is classified as a low fluctuation region. After completing the regional division of all monitoring points, the number and distribution of monitoring points in each region are counted. For example, the high fluctuation region contains 5 monitoring points, mainly concentrated at the reactor inlet; the medium fluctuation region contains 10 monitoring points, evenly distributed along the pipeline; and the low fluctuation region contains 8 monitoring points, located in the storage tank area. Finally, a monitoring point fluctuation distribution overview is generated, which includes the regional division results and monitoring point distribution information, including the fluctuation category of each monitoring point and its spatial distribution information.

[0067] The density filtering submodule calls the monitoring point fluctuation distribution overview, calculates the spatial density of monitoring points in the fluctuation area, filters dense areas, and obtains parameter distribution analysis results.

[0068] The spatial density of monitoring points in the fluctuation area is calculated using the following formula:

[0069]

[0070] Where ρ represents the spatial density of monitoring points in the fluctuation region, z i Let μ represent the parameter value of the i-th monitoring point, μ represent the mean of the parameter, σ represent the standard deviation of the parameter, p represent the weighting coefficient, A represent the area of ​​the monitoring region, and n represent the total number of monitoring points.

[0071] Spatial density of monitoring points in a fluctuating area refers to the distribution density or concentration of monitoring points within a specific area. Specifically, it represents the number of monitoring points set up per unit area or unit volume. In a fluctuating area, it refers to an area affected by dynamic changes or fluctuations. Monitoring points are used to track and record changes in relevant data. High spatial density of monitoring points can capture the fluctuation trends and changes within the area more precisely, thereby providing a more accurate basis for analysis and early warning.

[0072] The system receives an overview of the fluctuation distribution of monitoring points. This result includes the spatial location information of monitoring points within each fluctuation region and the fluctuation category to which they belong. For each fluctuation region (high fluctuation, medium fluctuation, low fluctuation), the spatial density ρ of monitoring points within it is calculated using the formula... z i The concentration value of sodium dihydrogen phosphate at the i-th monitoring point is obtained in real time by an online concentration sensor deployed on the production line. For example, if the concentration values ​​of three monitoring points in a high-fluctuation area are 98.2%, 97.9%, and 98.5% at a certain moment, then z1 = 98.2, z2 = 97.9, and z3 = 98.5. n is the total number of monitoring points in the fluctuation area, here n = 3. μ represents the average concentration of all monitoring points in the area over the past hour, assuming the average is 98.1%. σ represents the standard deviation of the concentration of all monitoring points in the area over the past hour, assuming the standard deviation is 0.2%. p represents the weighting coefficient, used to adjust the influence of dispersion on density, and is set to 2 based on experience. A represents the area of ​​the fluctuation area, which needs to be measured and set in advance according to the production line layout. Assuming the area of ​​the high-fluctuation area is 5 square meters.

[0073] The spatial density of monitoring points in this area is calculated as follows:

[0074] After calculating the spatial density of monitoring points in each fluctuation area, a density screening threshold is set, for example, 1.0. If the calculated spatial density of the fluctuation area is greater than this threshold, the area is considered a dense area and retained; otherwise, it is discarded. In this calculation, the density of the high fluctuation area is 1.05, which is greater than the threshold of 1.0, so it is screened as a dense area. Finally, the parameter distribution analysis results after density screening are obtained. These results will be used as input to the data hierarchical scheduling module. The parameter distribution analysis results identify the fluctuation areas with dense spatial distribution.

[0075] Specifically, such as Figure 2 , 4 As shown, the data hierarchical scheduling module includes:

[0076] The storage priority setting submodule calls the parameter distribution analysis results, sets the storage priority based on the parameter fluctuation area, sorts them according to the degree of fluctuation, and obtains the storage priority data of the monitoring points.

[0077] The system receives parameter distribution analysis results, which indicate different fluctuation regions (high fluctuation, medium fluctuation, and low fluctuation) and the density of monitoring point distribution within these regions. Data storage priorities are set based on the characteristics of the parameter fluctuation regions; regions with greater parameter fluctuations have higher data storage priorities. Specifically, the priority settings are: P1 (highest) for monitoring points in high fluctuation regions, P2 (medium) for monitoring points in medium fluctuation regions, and P3 (lowest) for monitoring points in low fluctuation regions. Within the same fluctuation region, if the region is identified as a dense region, the storage priority of all monitoring points within that region is increased by one level. For example, if a high fluctuation region is identified as a dense region, the storage priority of all monitoring points within that region is increased from P1 to P0 (very high). Through this setting, the storage priority data for each monitoring point is obtained, reflecting the importance of data storage for each monitoring point.

[0078] The storage frequency adjustment submodule calls the storage priority data of the monitoring points, adjusts the storage frequency according to the priority, sets high frequency storage for high fluctuation areas, medium frequency storage for medium fluctuation areas, and low frequency storage for low fluctuation areas, analyzes the changes in storage frequency, adjusts the size of the storage dataset, and obtains the data storage frequency partition dataset.

[0079] The system receives monitoring point storage priority data and adjusts the data storage frequency according to different priorities. The specific adjustment strategy is as follows: for monitoring points with priority P0 (ultra-high priority), data is stored once per second; for monitoring points with priority P1 (high priority), data is stored once every 5 seconds; for monitoring points with priority P2 (medium priority), data is stored once every 15 seconds; and for monitoring points with priority P3 (low priority), data is stored once every 30 seconds. For example, if monitoring point E has a storage priority of P1, its data storage frequency is adjusted to once every 5 seconds; if monitoring point F has a storage priority of P3, its data storage frequency is adjusted to once every 30 seconds. By adjusting the storage frequency, the impact on the overall storage dataset size is analyzed. Monitoring points with high-frequency storage generate more data, while monitoring points with low-frequency storage generate less data, thus forming a storage dataset partitioned according to fluctuation characteristics and priorities. Finally, a data storage frequency partitioned dataset is obtained. This dataset contains each monitoring point and its corresponding data storage frequency, which will serve as input to the anomaly linkage identification module. The data storage frequency partitioned dataset clearly defines the data storage frequency of each monitoring point.

[0080] Specifically, such as Figure 2 , 5 As shown, the anomaly linkage identification module includes:

[0081] The manufacturing feature recognition submodule extracts temperature, pressure, and flow rate from monitoring points based on the data storage frequency partition dataset to obtain a set of manufacturing feature parameters.

[0082] Based on the received data storage frequency partitioned dataset, which contains time-series data of manufacturing parameters such as temperature, pressure, and flow rate stored at different frequencies by various monitoring points, the average temperature, maximum pressure, and flow rate fluctuation (difference between maximum and minimum values) over the past 10 minutes are extracted from the data of each monitoring point to form the manufacturing characteristic parameter set for that monitoring point. For example, for monitoring point G, if the temperature data collected in the past 10 minutes is {30.1, 30.2, 30.3, 30.2, 30.1}℃, then its average temperature is 30.18℃, and the collected pressure data is {2.0, 2.1, 2.2, 2.1, ...}. If the pressure is 2.0 MPa, then the maximum pressure is 2.2 MPa. The collected flow rate data is {10.5, 10.3, 10.7, 10.4, 10.6} L / min, then the flow rate fluctuation range is 10.7 - 10.3 = 0.4 L / min. Therefore, the manufacturing characteristic parameter set of monitoring point G is {average temperature: 30.18℃, maximum pressure: 2.2 MPa, flow rate fluctuation range: 0.4 L / min}. Perform the same operation on all monitoring points to obtain the manufacturing characteristic parameter set containing all monitoring points. The data will be used for subsequent monitoring point classification, including the average temperature, maximum pressure, and flow rate fluctuation range of each monitoring point.

[0083] The monitoring point classification submodule calls the manufacturing feature parameter set, analyzes the differences in manufacturing features based on temperature, pressure, and flow rate, filters monitoring points with similar feature values, determines the category to which the monitoring points belong, divides the same manufacturing feature units, optimizes the classification boundary, and obtains the distribution of manufacturing feature units.

[0084] The system receives a set of manufacturing characteristic parameters, which includes the average temperature, maximum pressure, and flow rate fluctuation for each monitoring point. The monitoring points are classified according to set characteristic difference thresholds. For example, the threshold for average temperature difference is set to 0.5℃, the threshold for maximum pressure difference is set to 0.1MPa, and the threshold for flow rate fluctuation difference is set to 0.2L / min. For any two monitoring points, if the difference in average temperature is less than 0.5℃, the difference in maximum pressure is less than 0.1MPa, and the difference in flow rate fluctuation is less than 0.2L / min, then the manufacturing characteristics are considered similar, and the monitoring points with similar characteristic values ​​are grouped into the same category, i.e., the same manufacturing characteristic unit. For example, the characteristic parameters of monitoring point H are {30.2℃, 2.1MPa, 0.3L / min}, the characteristic parameters of monitoring point I are {30.4℃, 2.15MPa, 0.25L / min}, and the characteristic parameters of monitoring point J are {31... Comparing H and I (0.0℃, 2.0MPa, 0.1L / min), the temperature difference is 0.2℃ < 0.5℃, the pressure difference is 0.05MPa < 0.1MPa, and the flow rate difference is 0.05L / min < 0.2L / min. Therefore, H and I belong to the same manufacturing characteristic unit. Comparing H and J, the temperature difference is 0.8℃ > 0.5℃, so they do not belong to the same unit. In this way, all monitoring points are divided into different manufacturing characteristic units. The classification boundary is fine-tuned according to the actual distribution of monitoring points. For example, if the monitoring points in a certain unit are found to have obvious spatial clustering, the classification is further optimized to ensure that the monitoring points in the same manufacturing characteristic unit have a high degree of consistency in manufacturing characteristics. Finally, the distribution of manufacturing characteristic units is obtained. This distribution information will be passed to the linkage comparison generation submodule. The distribution of manufacturing characteristic units clarifies the set of monitoring points with similar manufacturing characteristics.

[0085] The linkage comparison generation submodule calls the manufacturing feature unit distribution, filters the monitoring points within the same unit, and obtains the abnormal linkage comparison relationship;

[0086] The system receives a distribution of manufacturing feature units, indicating which monitoring points belong to the same manufacturing feature unit. Within each identical manufacturing feature unit, any two or more monitoring points are considered to have a potential abnormal linkage relationship. For example, if manufacturing feature unit M contains monitoring points K, L, and N, then the monitoring point pairs (K, L), (K, N), (L, N), and (K, L, N) form an abnormal linkage control relationship. When an anomaly occurs at a monitoring point within the unit, the system checks whether other monitoring points within the same unit also exhibit anomalies. This is achieved by filtering all monitoring points within the same manufacturing feature unit. There are monitoring point combinations. For example, if there are three manufacturing feature units, unit 1 contains {A, B}, unit 2 contains {C, D, E}, and unit 3 contains {F}, then the generated abnormal linkage comparison relationships include {(A, B)}, {(C, D), (C, E), (D, E), (C, D, E)}. Note that unit 3 has only one monitoring point, so there is no linkage comparison relationship. Finally, the abnormal linkage comparison relationship is obtained, and this relationship will be passed to the trend synchronization mapping module. The abnormal linkage comparison relationship clarifies the potential linkage relationship between monitoring points within the same manufacturing feature unit.

[0087] Specifically, such as Figure 2 , 6 As shown, the trend synchronization mapping module includes:

[0088] The monitoring point trend analysis submodule calls the abnormal linkage comparison relationship, analyzes the fluctuation trend of manufacturing parameters, extracts time series data, and obtains the trend change characteristics of monitoring points;

[0089] The system receives abnormal linkage comparison relationships. For each linked monitoring point pair or group, it extracts the time series data of manufacturing parameters (such as temperature and pressure) fluctuations over the past 30 minutes from the historical database. For example, for a linkage relationship (monitoring point P, monitoring point Q), it extracts the temperature time series data over the past 30 minutes. The temperature data for monitoring point P is {30.1, 30.2, 30.3, 30.2, ..., 31.0}, and the temperature data for monitoring point Q is {35.1, 35.2, 35.3, 35.2, ..., 36.0}. It analyzes the time series data for each monitoring point, such as calculating its linear trend and fluctuation frequency, to obtain the trend change characteristics of each monitoring point. These characteristics will be used for subsequent data correlation judgment. Obtaining the trend change characteristics of the monitoring points describes the parameter fluctuation trend of the linked monitoring points over a period of time.

[0090] The data correlation judgment submodule calls the trend change characteristics of monitoring points, analyzes the consistency of the fluctuation trend of manufacturing parameters, calculates the correlation of monitoring point data, sets the correlation threshold, filters the correlation of monitoring point data, removes low correlation monitoring points, optimizes data matching relationship, and obtains the correlation filtering results of monitoring points.

[0091] The system receives trend change characteristics of monitoring points and analyzes the consistency of manufacturing parameter fluctuation trends for each pair or group of monitoring points with linkage relationships. For example, for monitoring points P and Q, temperature trend change characteristics are obtained respectively. The consistency of trends is quantified by calculating the Pearson correlation coefficient of temperature time series data. Assuming that the calculated temperature correlation coefficient between monitoring points P and Q is 0.85, and the correlation threshold is set to 0.7, if the correlation coefficient between two monitoring points is greater than or equal to this threshold, the data is considered to have a high correlation; otherwise, the correlation is considered to be low. For example, if the pressure correlation coefficient between monitoring points R and S is 0.5, which is lower than the threshold of 0.7, the linkage relationship between this pair of monitoring points is eliminated. After screening, the data matching relationship is optimized, and only monitoring point pairs or groups with high correlation are retained. Finally, the monitoring point association screening results are obtained, and these results are passed to the stability trend screening submodule. The monitoring point association screening results identify linked monitoring points with highly consistent fluctuation trends.

[0092] The stability trend filtering submodule calls the monitoring point association filtering results, filters the trend change data of the associated monitoring points, identifies the magnitude of production trend changes, and obtains the production stability change trend.

[0093] The magnitude of changes in production trends is determined using the following formula:

[0094]

[0095] Where ΔT represents the magnitude of change in production trend, T j T represents the production trend value at time j. j-1 α represents the production trend value at time j-1, α represents the weighting coefficient of the trend change magnitude, and m represents the total number of time periods.

[0096] The magnitude of production trend changes refers to the degree and magnitude of fluctuations in various indicators or output during the production process within a certain time period. It reflects the range of fluctuations in various factors in production activities (such as production efficiency, raw material supply, equipment operation status, etc.). A large magnitude of change usually means that the production process is unstable and affected by multiple factors, while a small magnitude of change indicates that production is relatively stable and less volatile. By monitoring the magnitude of production trend changes, companies can identify problems in a timely manner and take measures to optimize production processes and improve production efficiency.

[0097] Receive the monitoring point correlation filtering results, and for each highly correlated monitoring point pair or group, filter the trend change data over the past 15 minutes, identify the magnitude of the production trend change ΔT, and apply it to the formula. In the middle, T jThis represents the production trend value at time j. The production trend value can be the average or weighted value of key parameters such as temperature, pressure, or flow rate. For example, for associated monitoring points U and V, temperature values ​​are recorded every minute over the past 15 minutes, resulting in a time series {32.1, 32.2, 32.0, 32.3, ..., 32.5}. Here, we assume m = 15, T1 = 32.1, T2 = 32.2, ..., T... 15 =32.5, where α represents the weighting coefficient for the magnitude of trend change, used to adjust for the impact of the degree of change, and is set to 1.5 based on experience;

[0098] The magnitude of the change in production trend is calculated as follows:

[0099]

[0100] Assuming ΔT = 0.1 is calculated, the production trend change amplitude of all associated monitoring point groups is calculated to identify the production stability change trend. A higher ΔT value indicates that the production trend fluctuates greatly and has poor stability, while a lower ΔT value indicates that the production trend is relatively stable. Finally, the production stability change trend is obtained. This trend information will be used as input to the storage partition optimization module, reflecting the current stability level of the production process.

[0101] Specifically, such as Figure 2 , 7 As shown, the storage partition optimization module includes:

[0102] The spatial layout extraction submodule calls the production stability change trend, extracts the spatial layout data of monitoring points, identifies the corresponding location of monitoring points, and obtains the spatial layout information of monitoring points.

[0103] The system receives trends in production stability and simultaneously extracts spatial layout data corresponding to highly correlated monitoring points from a pre-stored database of factory equipment layout information. This data includes the three-dimensional coordinates of each monitoring point and the identifier of the equipment unit it is connected to. For example, monitoring point U has coordinates (10, 20, 5) and is connected to reactor A, while monitoring point V has coordinates (12, 21, 6) and is connected to the agitator of reactor A. Through the spatial layout data, the system identifies the specific location and physical connection relationship of each monitoring point on the production line and obtains the spatial layout information of the monitoring points. This information is then passed to the storage partitioning submodule. The spatial layout information of the monitoring points describes the location and connection relationship of each monitoring point in physical space.

[0104] The storage partitioning submodule calls the spatial layout information of the monitoring points, identifies layout features based on the corresponding locations of the monitoring points, filters dense areas of monitoring points, divides storage partitions, sets differentiated storage partition boundaries, adjusts the distribution balance of monitoring points, and obtains the storage partitioning results.

[0105] The system receives spatial layout information of monitoring points and identifies layout features based on the location of each monitoring point. For example, it determines whether there are areas where monitoring points are densely distributed in physical space, or whether there are multiple monitoring points connected to the same key equipment unit. It then filters out areas where monitoring points are relatively dense. For instance, if the spatial distance between monitoring points U, V, and W around reactor A is less than 2 meters, this area is designated as a potential storage partition. Based on the identified layout features, the system initially divides the storage partitions and sets differentiated storage partition boundaries. For example, for dense monitoring points around reactor A, a circular area centered on the reactor is defined as the storage partition boundary. Considering data balance, if the number of monitoring points in a partition is too large or too small, the partition boundary needs to be adjusted. For instance, if the number of monitoring points in a partition far exceeds the number of partitions, the boundary of that partition is appropriately reduced or it is merged with another partition. Finally, the system obtains the storage partition division result, which is then passed to the partition structure adjustment submodule. The storage partition division result defines the initial storage partitions and the monitoring points they contain.

[0106] The partition structure adjustment submodule calls the storage partitioning results, adjusts the storage partition structure, optimizes the data mapping of monitoring points, and obtains a secure storage solution for intelligent manufacturing monitoring data.

[0107] The system receives the storage partitioning results and adjusts the mapping relationship of the monitoring point storage data in the storage according to the partitioned storage. The monitoring point data in the same storage partition is stored in adjacent or the same physical storage area to optimize the access efficiency and security of the monitoring point storage data. For example, if partition 1 contains data of monitoring points U, V, and W, the data is mapped to the same disk array or logical volume in the storage. By adjusting the storage partition structure, centralized management and fast access of data are achieved. Through physical isolation and centralized management, the security and maintainability of data are improved, and a secure storage solution for intelligent manufacturing monitoring data is obtained. This solution defines the final storage partition structure and the mapping method of monitoring point data.

[0108] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate, characterized in that, The system includes: Based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, the parameter analysis module calculates the fluctuation range and frequency of manufacturing parameters, divides the parameters into three categories: high, medium, and low fluctuation, analyzes the distribution density of monitoring points, filters dense areas, and obtains the parameter distribution analysis results. Based on the parameter distribution analysis results, the data hierarchical scheduling module sets the data storage priority, adjusts the data storage frequency of differentiated fluctuation areas, sets high fluctuation areas as high frequency storage, medium fluctuation areas as medium frequency storage, and low fluctuation areas as low frequency storage, and obtains a data storage frequency partition dataset. The abnormal linkage identification module calls the data storage frequency partition dataset, extracts manufacturing feature parameters, filters monitoring points in the same type of manufacturing unit based on feature differences, and obtains abnormal linkage comparison relationships. Based on the aforementioned abnormal linkage comparison relationship, the trend synchronization mapping module analyzes the consistency of the fluctuation trend of manufacturing parameters, judges the correlation of monitoring point data, sets a correlation threshold, filters related monitoring points, and obtains the trend of production stability changes.

2. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 1, characterized in that: The parameter distribution analysis results include high fluctuation regions, medium fluctuation regions, and low fluctuation regions. The data storage frequency partition dataset includes high frequency storage regions, medium frequency storage regions, and low frequency storage regions. The abnormal linkage comparison relationship includes monitoring point units with similar temperature, monitoring point units with similar pressure, and monitoring point units with similar flow. The production stability change trend includes manufacturing parameter fluctuation trend, monitoring point consistency trend, correlation threshold, and correlated monitoring points.

3. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 1, characterized in that: The parameter distribution analysis results include: The fluctuation range calculation submodule measures and records manufacturing parameter values ​​based on the spatial distribution characteristics of sodium dihydrogen phosphate at the monitoring points, analyzes the fluctuation range and frequency of change between each monitoring point, and obtains parameter fluctuation values ​​based on the fluctuation range and frequency of change. The fluctuation classification submodule calls the parameter fluctuation value, identifies the fluctuation category based on the fluctuation characteristics of adjacent monitoring points, determines the fluctuation range to which the monitoring point belongs, sets the fluctuation classification standard, divides the monitoring points into three categories of areas: high fluctuation, medium fluctuation, and low fluctuation, analyzes the distribution of monitoring points in the area, and generates a monitoring point fluctuation distribution overview. The density filtering submodule calls the monitoring point fluctuation distribution overview, calculates the spatial density of monitoring points in the fluctuation area, filters dense areas, and obtains parameter distribution analysis results.

4. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 3, characterized in that: The spatial density of monitoring points in the fluctuation region is calculated using the following formula: Where ρ represents the spatial density of monitoring points in the fluctuation region, z i Let μ represent the parameter value of the i-th monitoring point, μ represent the mean of the parameter, σ represent the standard deviation of the parameter, p represent the weighting coefficient, A represent the area of ​​the monitoring region, and n represent the total number of monitoring points.

5. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 3, characterized in that: The data hierarchical scheduling module includes: The storage priority setting submodule calls the parameter distribution analysis results, sets the storage priority according to the parameter fluctuation area, sorts them according to the degree of fluctuation, and obtains the monitoring point storage priority data; The storage frequency adjustment submodule calls the storage priority data of the monitoring points, adjusts the storage frequency according to the priority, sets high frequency storage for high fluctuation areas, medium frequency storage for medium fluctuation areas, and low frequency storage for low fluctuation areas, analyzes the changes in storage frequency, adjusts the size of the storage dataset, and obtains the data storage frequency partition dataset.

6. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 5, characterized in that: The abnormal linkage identification module includes: The manufacturing feature recognition submodule extracts the temperature, pressure, and flow rate of the monitoring points based on the data storage frequency partition dataset to obtain a set of manufacturing feature parameters. The monitoring point classification submodule calls the manufacturing feature parameter set, analyzes the differences in manufacturing features based on temperature, pressure, and flow rate, filters monitoring points with similar feature values, determines the category to which the monitoring points belong, divides the same manufacturing feature units, optimizes the classification boundary, and obtains the distribution of manufacturing feature units. The linkage comparison generation submodule calls the manufacturing feature unit distribution, filters monitoring points within the same unit, and obtains abnormal linkage comparison relationships.

7. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 6, characterized in that: The trend synchronization mapping module includes: The monitoring point trend analysis submodule calls the abnormal linkage comparison relationship to analyze the fluctuation trend of manufacturing parameters, extract time series data, and obtain the trend change characteristics of monitoring points; The data correlation judgment submodule calls the trend change characteristics of the monitoring points, analyzes the consistency of the fluctuation trend of manufacturing parameters, calculates the data correlation of the monitoring points, sets the correlation threshold, filters the data correlation of the monitoring points, removes low correlation monitoring points, optimizes the data matching relationship, and obtains the monitoring point correlation filtering results. The stability trend filtering submodule calls the monitoring point association filtering results, filters the trend change data of the associated monitoring points, identifies the magnitude of production trend changes, and obtains the production stability change trend.

8. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 7, characterized in that: The magnitude of the change in the production trend is determined using the following formula: Where ΔT represents the magnitude of change in production trend, T j T represents the production trend value at time j. j-1 α represents the production trend value at time j-1, α represents the weighting coefficient of the trend change magnitude, and m represents the total number of time periods.

9. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 1, characterized in that: The system also includes a storage partition optimization module: Based on the trend of production stability changes, the storage partition optimization module extracts the spatial layout information of monitoring points, identifies the corresponding positions between monitoring points, divides storage partitions according to layout characteristics, adjusts the storage partition structure, and obtains a secure storage solution for intelligent manufacturing monitoring data. The intelligent manufacturing monitoring data security storage solution includes storage partitioning, storage partition structure adjustment, extraction of monitoring point spatial layout information, and identification of corresponding monitoring point locations.

10. The IoT-based intelligent manufacturing monitoring system for sodium dihydrogen phosphate according to claim 9, characterized in that: The storage partition optimization module includes: The spatial layout extraction submodule calls the production stability change trend, extracts the spatial layout data of the monitoring points, identifies the corresponding location of the monitoring points, and obtains the spatial layout information of the monitoring points. The storage partitioning submodule calls the spatial layout information of the monitoring points, identifies layout features based on the corresponding locations of the monitoring points, filters dense areas of monitoring points, divides storage partitions, sets differentiated storage partition boundaries, adjusts the distribution balance of monitoring points, and obtains the storage partitioning results. The partition structure adjustment submodule calls the storage partitioning results, adjusts the storage partition structure, optimizes the data mapping of monitoring points, and obtains a secure storage solution for intelligent manufacturing monitoring data.

Images