CN114118982B - Production control system combining cloud computing and edge computing - Google Patents
Production control system combining cloud computing and edge computing Download PDFInfo
- Publication number
- CN114118982B CN114118982B CN202111501614.4A CN202111501614A CN114118982B CN 114118982 B CN114118982 B CN 114118982B CN 202111501614 A CN202111501614 A CN 202111501614A CN 114118982 B CN114118982 B CN 114118982B
- Authority
- CN
- China
- Prior art keywords
- production
- data
- monitoring
- information
- analysis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 250
- 238000012544 monitoring process Methods 0.000 claims abstract description 126
- 238000004458 analytical method Methods 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 45
- 230000008569 process Effects 0.000 claims abstract description 26
- 230000006978 adaptation Effects 0.000 claims abstract description 23
- 238000013515 script Methods 0.000 claims abstract description 6
- 238000002347 injection Methods 0.000 claims abstract description 5
- 239000007924 injection Substances 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 19
- 239000011159 matrix material Substances 0.000 claims description 16
- 230000002159 abnormal effect Effects 0.000 claims description 13
- 239000000463 material Substances 0.000 claims description 6
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 238000003754 machining Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 238000001816 cooling Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 3
- 230000005856 abnormality Effects 0.000 claims description 2
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000010219 correlation analysis Methods 0.000 claims description 2
- 238000005520 cutting process Methods 0.000 claims description 2
- 238000007405 data analysis Methods 0.000 claims description 2
- 239000010705 motor oil Substances 0.000 claims description 2
- 238000004891 communication Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000008713 feedback mechanism Effects 0.000 description 2
- 238000000195 production control method Methods 0.000 description 2
- 101100295776 Drosophila melanogaster onecut gene Proteins 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000011022 operating instruction Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2474—Sequence data queries, e.g. querying versioned data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/248—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Probability & Statistics with Applications (AREA)
- Fuzzy Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- General Factory Administration (AREA)
Abstract
A cloud computing and edge computing combined production control system, comprising: the system comprises an instruction analysis module, a monitoring configuration updating module, an information adaptation module, a production execution module and a production analysis module, wherein the instruction analysis module receives a production control instruction from a cloud service platform and analyzes content to obtain a production task, an operation instruction and a process specification; the monitoring configuration updating module receives monitoring configuration information used for adjusting and describing a production monitoring target from the cloud service platform, analyzes the checking configuration information, then monitors and configures and verifies the effectiveness and applicable equipment, and completes generation and injection of related execution scripts according to the configuration information; the information adaptation module performs semantic to information protocol consistency adaptation on heterogeneous equipment in the production unit; the production execution module is connected with the edge end and the equipment end, and acquires and transmits an operation instruction, a production state and production data; the production analysis module is used for analyzing and feeding back the production process of the production unit. According to the cloud-side-end monitoring system and method, edge-end monitoring of a production process can be achieved, and cloud-side-end effective production control coordination is achieved by matching with a cloud service platform.
Description
Technical Field
The invention relates to a technology in the field of cloud computing, in particular to a production control system combining cloud computing and edge computing.
Background
Most of the existing production monitoring modes are cloud-end type centralized unified monitoring, namely monitoring of the production process is completed by a cloud service platform, and the existing production monitoring modes have the limitations of monitoring lag and large communication pressure and various error hidden dangers. Meanwhile, a single-device 'one-tool' type early warning mode is still adopted for the early warning of the time sequence data, so that the effective utilization of early warning information is not facilitated. By integrating the background, a production control system combining cloud computing and edge computing is designed, an analysis processing module for a business process is added in an edge end, more decision authorities are given to edge nodes, classified and graded production feedback is achieved by optimizing a time sequence data prediction processing method, and a production control mode combining cloud-edge-end is realized.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a production control system combining cloud computing and edge computing, which constructs an edge terminal based on a production manufacturing unit, receives and transmits a production instruction from a cloud service platform, and gives authority for edge terminal production monitoring and early warning, so that the production control system has better autonomy. And a prediction compensation and dynamic monitoring threshold value is provided for the prediction of time sequence data, so that a prediction result is more accurate. Therefore, edge monitoring of the production process is achieved, and cloud-edge-end effective production control coordination is achieved by matching with a cloud service platform.
The invention is realized by the following technical scheme:
The invention relates to a production control system combining cloud computing and edge computing, which comprises: the system comprises an instruction analysis module, a monitoring configuration updating module, an information adaptation module, a production execution module and a production analysis module, wherein: the instruction analysis module receives a production control instruction from the cloud service platform and analyzes the content to obtain a production task, an operation instruction and a process specification; the monitoring configuration updating module receives monitoring configuration information used for adjusting and describing a production monitoring target from the cloud service platform, analyzes the monitoring configuration information, verifies the validity and applicable equipment of the monitoring configuration information, and completes generation and injection of related execution scripts according to the configuration information; the information adaptation module performs semantic to information protocol consistency adaptation on heterogeneous equipment in the production unit; the production execution module is connected with the edge end and the equipment end, and acquires and transmits an operation instruction, a production state and production data; the production analysis module is used for analyzing and feeding back the production process of the production unit.
The cloud service platform is positioned at the upper layer of the system, is a starting point of production task execution and an end point of production monitoring in the system, transmits a production instruction and a JSON file of monitoring configuration through a service call RESTful interface, and requests production feedback information JSON file of the edge monitoring system.
The production control instruction comprises: production task information, equipment operating instructions, process specifications, and base information. The production task information is used for recording basic content information of the secondary production task, the equipment operation instruction records specific operation of related production equipment which needs to be completed in the secondary production task, the process specification defines process specification information in production, and the basic information records other information related to production.
The monitoring configuration information comprises: monitoring an index list, an index calculation rule and an index threshold list. The monitoring index list lists the target indexes to be monitored currently, the index calculation rule defines a calculation specific rule of the indexes to be calculated, and the index threshold list specifically lists the acceptable thresholds of the monitoring indexes.
The information adaptation module comprises: a semantic processing unit and a protocol processing unit, wherein: the semantic processing unit is used for assisting in issuing production control instructions and acquiring monitoring indexes by constructing domain knowledge graphs, providing word sense disambiguation and semantic matching functions, and defining the relation between instruction information and equipment reality, so that the method is a powerful support for ensuring effective execution and smooth monitoring and expansion of the instructions; the protocol processing unit helps the edge end system to effectively manage and control each device of the production unit through protocol adaptation and protocol conversion under the condition that the protocol difference between the service systems of each device in the production unit is large and the production instruction cannot be directly received.
The production analysis module comprises: the system comprises a monitoring data acquisition unit, a state data monitoring unit, a time sequence data early warning unit and a production database, wherein: the monitoring data acquisition unit comprises monitoring information analysis and data preprocessing, wherein the monitoring information analysis is used for analyzing and processing the collected and acquired monitoring information, the data preprocessing is used for carrying out screening, classifying and arranging operations on the monitoring data in sequence, receiving the time-series data in a sliding window mode and completing all processing operations before analysis and early warning; the status data monitoring unit will analyze status type data in the production monitoring data. The state data refer to non-time sequence data such as the production flow rate, and the monitoring analysis of the data comprises three steps of historical data statistics, state characteristic value calculation and abnormal state monitoring. Wherein the historical data statistics are obtained by acquiring historical production data and statistically calculating relevant production state data thereof so as to compare and check the evolution of the production state time dimension. The state eigenvalue calculation will calculate the relevant state data in the current production task. Abnormal state checking, namely comparing historical production state data with current production state data, current production state data with an index threshold value, and recording the abnormal data in the abnormal state checking into a production abnormal report; the time sequence data prediction unit analyzes time sequence type data in the production monitoring data. The time sequence data refer to time-frequency data which are continuously transmitted, such as engine oil temperature, and the monitoring analysis of the data comprises three steps of time sequence data prediction, prediction data compensation and dynamic threshold analysis. The time sequence data prediction is to predict the trend of the current data through a certain time sequence prediction algorithm and model. The prediction data compensation refers to analyzing the association relation of each link in production, and compensating and adjusting the prediction data in the subsequent links. The dynamic threshold analysis refers to the analysis and calculation of a dynamic monitoring threshold by considering the actual fluctuation condition of data between similar orders, and the monitoring, early warning and reminding are carried out by matching with a set threshold; the production database is used for storing and recording production tasks, execution processes, execution results and production feedback information of past times and providing data support for production data analysis.
The production feedback information comprises a production abnormality report and a monitoring early warning prompt, and is used for subsequent analysis and adjustment of the production process by the cloud service platform.
The prediction data compensation and dynamic threshold analysis specifically comprise: the production manufacturing unit comprises multiple kinds of equipment with compact production association, and the execution of the preceding links in the production execution can cause certain influence on the execution of the following links, for example, in a compact machining process, the cooling gun vortex tube has too low flow speed, so that the temperature of a cutter is not reduced in time, the cutting edge is deformed, and the machining precision is deviated. Therefore, the time sequence data change of the previous production link often has a certain association relation with the monitoring data of the subsequent production link. In such cases, the prediction of the time series data of a certain device in a unit directly and individually may cause deviation of the prediction result. Therefore, the present invention contemplates compensation adjustments to the predicted values in the subsequent production link by using the predicted compensation ΔΔd. The method comprises the following steps:
1) The production process comprises the steps of setting a preamble production link device P a and a direct subsequent production link device P b, wherein a is not equal to b, and obtaining time sequence data sets related to each other in two production links respectively from production practice and expert experience Then there is an association matrix of X a and X b Wherein the method comprises the steps of Representing dataAndDegree of association between the two.
2) Taking the values of m sliding windows before the current production record from a production database, and obtaining the product through gray scale correlation analysis and calculationWhere k represents information recorded in the kth sliding window, The resolution coefficient ρ is used to weaken the influence of abnormal values on the correlation space, so as to better embody the system integrity, and the value is 0.5.x a (k) denotes the device P a in the kth recordingSpecific values of (2).
3) The resulting time series data predictors for device P a and device P b, respectively, are based on algorithms including, but not limited to, an integrated autoregressive moving average (ARIMA) prediction algorithm
4) Combining the predicted value matrix D a with the associated matrix R a,b to complete the prediction compensation of D b and obtain the final predicted result of the time sequence data of the equipment P b
Considering that the time sequence data monitoring threshold is usually a manually set critical value, the time sequence data monitoring threshold has certain difference with the data change range in actual production, and the early warning application range is limited. In view of the problem, the invention provides a dynamic early warning threshold method based on production similarity, which realizes hierarchical early warning of time sequence data by combining a dynamic threshold and a critical value. The method comprises the following specific steps:
1) According to the monitoring time sequence data x i of the equipment P i, each sliding window n is included in one historical order record, and the average value of x i in the kth window is recorded as Calculating the comprehensive mean value of each window
2) The influence of different production order contents on the execution of production unit equipment comes from two parts of material selection and process specification, and the material characteristic information, the characteristic processing steps and the characteristic process requirements are coded and described based on a system unified process coding method to obtain a material characteristic information matrix MC, a characteristic processing step matrix PS and a characteristic process requirement matrix PR, wherein the order O i can be expressed as O i = |MC PS PR|. By expanding the matrix into a high-dimensional vector, the similarity s a,b between the two orders can be calculated using a calculation method that includes, but is not limited to, cosine similarity.
3) By performing similarity calculation on the first m production order records in the production database and the current production order, a similarity matrix s= |s 1 s2 … sm | of m records can be obtained, where S i represents the similarity between the order i and the current order.
4) According to the calculated comprehensive average value of the first m ordersSimilarity with each order and the current order, taking the latter as weight, obtaining the dynamic threshold value of the time sequence data x i through weighted averageWhere λ is the weight and represents the degree of deviation between the dynamic threshold and the weighted mean.
5) And (5) data early warning. The data early warning is divided into three stages, and the severity is respectively primary early warning, secondary early warning and tertiary early warning from high to low, and specifically comprises the following steps: when the predicted value D' i exceeds the critical value mu, performing primary early warning; when the predicted value D' i is in the interval [ (V d +mu)/2 mu ], the second-level early warning is carried out; and when the predicted D' i is in the interval V d,(Vd +mu)/2), carrying out three-level early warning.
Technical effects
Compared with the existing production process monitoring technology, the cloud-side-end cooperation monitoring system reduces the monitoring pressure of the cloud service platform on the production process, and achieves the production monitoring by constructing the edge-end monitoring system based on the production unit, so that the production monitoring dimension is expanded; the semantic matching of heterogeneous equipment of the production unit is completed through an adaptation module based on the domain knowledge graph, so that the effective issuing of production instructions and the acquisition of monitoring data are ensured; meanwhile, the state data and the time sequence data in the production process data are respectively processed and analyzed, and the classification and grading feedback of the production monitoring is completed by establishing the time sequence data prediction compensation and the dynamic monitoring threshold value. Under the industrial manufacturing background of cloud-edge cooperation, the problems of monitoring lag and low early warning information practicality in the existing production process monitoring are integrally solved.
Drawings
FIG. 1 is a schematic diagram of a system according to the present invention;
fig. 2 is a flow chart of an embodiment.
Detailed Description
As shown in fig. 1, this embodiment relates to a production control system combining cloud computing and edge computing, including: the system comprises an instruction analysis module, a monitoring configuration updating module, an information adaptation module, a production execution module and a production analysis module, wherein: the instruction analysis module receives a production control instruction from the cloud service platform and analyzes the content to obtain a production task, an operation instruction and a process specification; the monitoring configuration updating module receives monitoring configuration information used for adjusting and describing a production monitoring target from the cloud service platform, analyzes the monitoring configuration information, verifies the validity and applicable equipment of the monitoring configuration information, and completes generation and injection of related execution scripts according to the configuration information; the information adaptation module performs semantic to information protocol consistency adaptation on heterogeneous equipment in the production unit; the production execution module is connected with the edge end and the equipment end, and acquires and transmits an operation instruction, a production state and production data; the production analysis module is used for analyzing and feeding back the production process of the production unit.
As shown in fig. 2, in the production control method based on the combination of cloud computing and edge computing of the system in this embodiment, the edge production control system receives the production control information and the monitoring configuration information sent by the cloud service platform as the start point of the execution of the service flow, completes the production analysis, returns the production feedback information including the production anomaly report and the monitoring early warning reminder to the cloud service platform as the end point of the execution of the service flow, and the production unit layer includes various internet of things devices governed by the production end member and serves as the data source of the edge monitoring system.
The production control method specifically comprises the following steps:
Step 1) an instruction analysis module receives a production instruction JSON file and a monitoring configuration information JSON file through a service call RESTful interface, completes analysis of contents in the production instruction JSON file, and calls an information adaptation module to further process the analyzed information, wherein: generating and injecting configuration scripts for the monitoring configuration file; the information adaptation module carries out heterogeneous equipment adaptation on the instruction information and the monitoring configuration, and a semantic adaptation module based on the domain knowledge graph is called in the semantic adaptation process to carry out some word sense disambiguation and semantic matching.
Step 2) after the analysis and adaptation are completed, the equipment operation instruction applicable to the corresponding internet of things equipment is sent to a production execution module; the production execution module conveys the operation instruction to the related object associated equipment, completes the specific issuing of the operation instruction, and carries out two-way communication with the production unit through an OPC server interface, and collects index information to be monitored; after the production unit related object-associated equipment acquires specific instruction operation, the production task starts to be executed, and production process data is continuously uploaded through an equipment new data acquisition OPC server interface; the production execution module which collects the data of the production process transmits the data in real time and invokes the production analysis module; the production analysis module finishes screening and classifying the production process data through the monitoring data acquisition unit, and sequentially arranges the time sequence data, then transmits the production state data to the state data detection unit, and transmits the production time sequence data to the time sequence data early warning unit.
And 3) the state data monitoring unit calculates the state characteristic value of the current production according to the index calculation method in the monitoring configuration information, and encapsulates the state index exceeding the threshold value into an abnormal report JSON file.
And 4) sequentially receiving the collected data according to the sliding window configuration by the time sequence data early warning unit, grouping and completing the data, finally predicting the data by establishing a time sequence prediction model, wherein the time sequence prediction model comprises but is not limited to a regression prediction model and a Kalman filtering prediction model, and then establishing an early warning dynamic threshold by analyzing the actual fluctuation condition of the data of the historical similar order.
And 5) after the data prediction is completed, compensating and adjusting the prediction result by analyzing the relevance between production links, carrying out grading early warning based on a dynamic threshold value and a set threshold value, and sending a monitoring early warning reminding JSON file to realize the production monitoring of the production unit by the edge end system and complete the production monitoring coordination of cloud-edge-end.
Table 1 comparison of technical characteristics
Compared with the prior art, in the process of executing production monitoring, the method and the device for monitoring the cloud service platform reduce the load of the cloud service platform by lowering the process monitoring authority to the edge end based on the production unit, so that the self-monitoring of the edge end is realized. In the production control instruction analysis stage, receiving and analyzing the management control instruction based on the production unit, avoiding a mode that a cloud service platform needs to set an independent instruction for each production device, and reducing cloud management control pressure; by constructing a monitoring configuration updating module, script updating injection is automatically completed when a monitoring rule changes, so that online updating of monitoring configuration is realized, while the prior art project is set in advance by manpower, and online adjustment of monitoring content cannot be realized; for the received data information, compared with the semantic analysis by means of manual mode, the semantic matching is carried out by constructing the domain knowledge graph, so that the error rate is less processed; in the prediction and early warning processing of time sequence data, the prior art project only predicts data for single equipment, and the early warning threshold value is set in advance and is difficult to adjust in time; in the management and control feedback output processing, compared with unified judgment by utilizing static monitoring rules, the method and the system for classifying and classifying the state data and the time sequence data by the aid of the classification and classification feedback mechanism are perfected, classification judgment is carried out on early warning of the time sequence data, and further management and control analysis of a cloud service platform is effectively supported.
From the aspect of management and control timeliness, the cloud-side-end hierarchical collaboration mode is beneficial, and compared with the cloud service platform for uniformly managing and controlling the production process of each Internet of things device, the cloud service platform for directly managing and controlling communication and processing pressure is reduced. By combining cloud computing and edge computing, monitoring analysis closer to a production end is beneficial to improving timeliness of anomaly discovery and early warning.
Compared with the method for directly calculating and predicting the production time sequence data of single equipment from the aspect of prediction accuracy, the method provided by the invention has the advantages that in the production analysis module, the degree of the associated influence of the preamble production link on the current link is calculated, the compensation adjustment is provided for the time sequence data prediction, and the prediction accuracy of the time sequence data is ensured.
From the aspect of early warning adaptability, the invention calculates the dynamic monitoring threshold value by carrying out statistical analysis on the actual fluctuation range of similar production order data, and avoids using the early warning threshold value to carry out 'one-cut' type threshold value control on different production conditions. And the method and the device realize a grading feedback strategy for abnormal early warning together with a set critical value, and improve the adaptability of monitoring early warning to different production conditions.
According to the method, the edge end based on the production unit is established, the management and control instruction from the cloud service platform is received and transmitted, the monitoring configuration information is analyzed, the monitoring content is updated on line, and the semantic analysis of the received data is realized based on the domain knowledge graph; the system performs classified monitoring and early warning on the production state data and the production time sequence data by acquiring and analyzing the execution process data from the production unit. And aiming at time sequence data, compensating and adjusting a predicted value by analyzing influence association among production links, analyzing and sorting historical monitoring data to obtain a dynamic monitoring threshold value, and realizing hierarchical prediction by matching with an early warning critical value to provide multi-angle information for further production analysis and decision making of a cloud service platform.
In the existing production control system, the cloud service platform directly monitors all production equipment and does not perform management and control layering. The relevance among production links is not considered in the prediction processing of the time series data, the prediction data is pre-warned by comparing with a unique pre-warning critical value, and certain limitation exists in processing. Compared with the prior art, the method and the device can better break through the limitation of the current mode, monitor and analyze at the edge end closer to production equipment, ensure the control timeliness, and simultaneously, enhance the feedback effect by improving the prediction accuracy and utilizing a grading feedback mechanism, so that a powerful support can be provided for the whole production control.
The foregoing embodiments may be partially modified in numerous ways by those skilled in the art without departing from the principles and spirit of the invention, the scope of which is defined in the claims and not by the foregoing embodiments, and all such implementations are within the scope of the invention.
Claims (4)
1. A production control system combining cloud computing and edge computing, comprising: the system comprises an instruction analysis module, a monitoring configuration updating module, an information adaptation module, a production execution module and a production analysis module, wherein: the instruction analysis module receives a production control instruction from the cloud service platform and analyzes the content to obtain a production task, an operation instruction and a process specification; the monitoring configuration updating module receives monitoring configuration information used for adjusting and describing a production monitoring target from the cloud service platform, analyzes the checking configuration information, then monitors and configures and verifies the effectiveness and applicable equipment, and completes generation and injection of related execution scripts according to the configuration information; the information adaptation module performs semantic to information protocol consistency adaptation on heterogeneous equipment in the production unit; the production execution module is connected with the edge end and the equipment end, and acquires and transmits an operation instruction, a production state and production data; the production analysis module is used for analyzing and feeding back the production process of the production unit;
The production control instruction comprises: production task information, equipment operation instructions, process specifications and basic information; the production task information is used for recording basic content information of the secondary production task, the equipment operation instruction records specific operation which needs to be completed in the secondary production task of related production equipment, the process specification defines process specification information in production, and the basic information records other information related to production;
The monitoring configuration information comprises: monitoring an index list, an index calculation rule and an index threshold list; the monitoring index list enumerates target indexes to be monitored currently, the index calculation rule defines a calculation specific rule of the indexes to be calculated, and the index threshold list specifically enumerates acceptable thresholds of the monitoring indexes;
The production analysis module comprises: the system comprises a monitoring data acquisition unit, a state data monitoring unit, a time sequence data early warning unit and a production database, wherein: the monitoring data acquisition unit comprises monitoring information analysis and data preprocessing, wherein the monitoring information analysis is used for analyzing and processing the collected and acquired monitoring information, the data preprocessing is used for screening, classifying and arranging the monitoring data in sequence, and receiving the time sequence data in a sliding window mode to complete all processing operations before analysis and early warning; the state data monitoring unit is used for analyzing state data in the production monitoring data; the state data refer to non-time sequence data such as the production flow rate, and the monitoring analysis of the data comprises three steps of historical data statistics, state characteristic value calculation and abnormal state monitoring; wherein the historical data statistics are obtained by acquiring historical production data and statistically calculating relevant production state data thereof so as to compare and check the evolution of the production state time dimension; the state characteristic value calculation is used for calculating relevant state data in the current production task; abnormal state checking, namely comparing historical production state data with current production state data, current production state data with an index threshold value, and recording the abnormal data in the abnormal state checking into a production abnormal report; the time sequence data prediction unit is used for analyzing time sequence type data in the production monitoring data; the time sequence data is time-frequency data which is continuously transmitted by the engine oil temperature, and the monitoring analysis of the data comprises three steps of time sequence data prediction, prediction data compensation and dynamic threshold analysis; the time sequence data prediction is to predict the trend of the current data through a certain time sequence prediction algorithm and model; the prediction data compensation refers to analyzing the association relation of each link in production, and compensating and adjusting the prediction data in the subsequent links; the dynamic threshold analysis refers to the analysis and calculation of a dynamic monitoring threshold by considering the actual fluctuation condition of data between similar orders, and the monitoring, early warning and reminding are carried out by matching with a set threshold; the production database is used for storing and recording production tasks, execution processes, execution results and production feedback information of past times and providing data support for production data analysis;
The production feedback information comprises a production abnormality report and a monitoring early warning prompt, and is used for subsequent analysis and adjustment of the production process by the cloud service platform;
The prediction data compensation and dynamic threshold analysis specifically comprise: the production manufacturing unit comprises a plurality of kinds of equipment with compact production association, the execution of a preceding link in the production execution can cause a certain influence on the execution of a following link, and in the compact machining process, the cooling gun vortex tube has too low flow speed to cause untimely cooling of a cutter, so that a cutting edge is deformed, and the machining precision is deviated; therefore, the time sequence data change of the previous production link often has a certain association relation with the monitoring data of the subsequent production link; in such cases, the single prediction of the time sequence data of a certain device in the unit can cause the deviation of the prediction result;
The predicted value in the subsequent production link is compensated and adjusted by using the predicted compensation delta D, which is specifically as follows:
1) The production process comprises the steps of setting a preamble production link device P a and a direct subsequent production link device P b, wherein a is not equal to b, and obtaining time sequence data sets related to each other in two production links respectively from production practice and expert experience Then there is an association matrix of X a and X b Wherein the method comprises the steps of Representing dataAndThe degree of association between the two;
2) Taking the values of m sliding windows before the current production record from a production database, and obtaining the product through gray scale correlation analysis and calculation Where k represents information recorded in the kth sliding window, The resolution factor ρ is used to attenuate the effect of outliers on the correlation space to better characterize the system integrity, x a (k) represents the device P a in the kth recordThe specific value of (3);
3) The time series data predicted values of the device P a and the device P b obtained by the comprehensive autoregressive moving average (ARIMA) prediction algorithm are respectively
4) Combining the predicted value matrix D a with the associated matrix R a,b to complete the prediction compensation of D b and obtain the final predicted result of the time sequence data of the equipment P b
2. The cloud computing and edge computing combined production control system according to claim 1, wherein the cloud service platform is located at an upper layer of the system, is a starting point of execution of a production task and an important point of production monitoring in the system, transmits a production instruction and a JSON file of monitoring configuration through a service call RESTful interface, and requests production feedback information JSON file of an edge monitoring system.
3. The cloud computing and edge computing combined production control system of claim 1, wherein the information adaptation module comprises: a semantic processing unit and a protocol processing unit, wherein: the semantic processing unit is used for assisting in issuing production control instructions and acquiring monitoring indexes by constructing domain knowledge graphs, providing word sense disambiguation and semantic matching functions, and defining the relation between instruction information and equipment reality, so that the method is a powerful support for ensuring effective execution and smooth monitoring and expansion of the instructions; the protocol processing unit helps the edge end system to effectively manage and control each device of the production unit through protocol adaptation and protocol conversion under the condition that the protocol difference between the service systems of each device in the production unit is large and the production instruction cannot be directly received.
4. A dynamic early warning threshold method based on the production similarity of the system of any one of claims 1 to 3, characterized in that the method realizes hierarchical early warning of time series data by combining a dynamic threshold and a critical value, and specifically comprises the following steps:
1) According to the monitoring time sequence data x i of the equipment P i, each sliding window n is included in one historical order record, and the average value of x i in the kth window is recorded as Calculating the comprehensive mean value of each window
2) The influence of different production order contents on the execution of production unit equipment comes from two parts of material selection and process specification, and based on a system unified process coding method, material characteristic information, characteristic processing steps and characteristic process requirements are coded and described to obtain a material characteristic information matrix MC, a characteristic processing step matrix PS and a characteristic process requirement matrix PR, wherein the order O i is expressed as O i = |MC PS PR|; by expanding the matrix into a high-dimensional vector and using a cosine similarity calculation method, the similarity s a,b between two orders can be calculated;
3) By performing similarity calculation on the records of the first m production orders in the production database and the current production order, a similarity matrix S= |s 1 s2…sm | of the records of m times can be obtained, wherein S i represents the similarity between the order i and the current order;
4) According to the comprehensive average value of the previous m orders Similarity with each order and the current order, taking the latter as weight, obtaining the dynamic threshold value of the time sequence data x i through weighted averageWherein λ is a weight representing the degree of deviation between the dynamic threshold and the weighted mean;
5) The three-level data early warning specifically comprises: when the predicted value D' i exceeds the critical value mu, performing primary early warning; when the predicted value D' i is in the interval [ (V d +mu)/2 mu ], the second-level early warning is carried out; and when the predicted D' i is in the interval V d,(Vd +mu)/2), carrying out three-level early warning.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111501614.4A CN114118982B (en) | 2021-12-09 | 2021-12-09 | Production control system combining cloud computing and edge computing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111501614.4A CN114118982B (en) | 2021-12-09 | 2021-12-09 | Production control system combining cloud computing and edge computing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114118982A CN114118982A (en) | 2022-03-01 |
| CN114118982B true CN114118982B (en) | 2024-07-26 |
Family
ID=80363880
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111501614.4A Active CN114118982B (en) | 2021-12-09 | 2021-12-09 | Production control system combining cloud computing and edge computing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114118982B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115665590B (en) * | 2022-10-21 | 2023-06-20 | 北京中电飞华通信有限公司 | Internet of things data acquisition system and method based on eSIM card and 5G communication |
| CN117048113B (en) * | 2023-10-13 | 2023-12-22 | 江苏爱箔乐铝箔制品有限公司 | Dynamic precision detection method and system for stamping aluminum foil cutlery box |
| CN118131681B (en) * | 2024-04-17 | 2024-09-13 | 北京铁力山科技股份有限公司 | Intelligent control platform based on artificial intelligence |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110768914A (en) * | 2019-10-23 | 2020-02-07 | 上海交通大学 | Decentralized Internet of things gateway system based on semantic scene instance migration |
| CN112379653A (en) * | 2020-12-01 | 2021-02-19 | 国能信控互联技术有限公司 | Smart power plant management and control system based on micro-service architecture |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10776194B2 (en) * | 2018-01-31 | 2020-09-15 | Splunk Inc. | Self-monitor for computing devices of a distributed computing system |
| CN113051830B (en) * | 2021-04-01 | 2023-04-18 | 重庆大学 | Intelligent production line dynamic error prediction system, control method and digital twin system |
| CN113255584B (en) * | 2021-06-22 | 2021-10-19 | 德明通讯（上海）股份有限公司 | Fault diagnosis and monitoring system based on edge calculation |
-
2021
- 2021-12-09 CN CN202111501614.4A patent/CN114118982B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110768914A (en) * | 2019-10-23 | 2020-02-07 | 上海交通大学 | Decentralized Internet of things gateway system based on semantic scene instance migration |
| CN112379653A (en) * | 2020-12-01 | 2021-02-19 | 国能信控互联技术有限公司 | Smart power plant management and control system based on micro-service architecture |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114118982A (en) | 2022-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114118982B (en) | Production control system combining cloud computing and edge computing | |
| US10902368B2 (en) | Intelligent decision synchronization in real time for both discrete and continuous process industries | |
| JP7009438B2 (en) | Computer systems and methods for monitoring key performance indicators (KPIs) using time series pattern models | |
| US20220024607A1 (en) | Predictive maintenance model design system | |
| US7738975B2 (en) | Analytical server integrated in a process control network | |
| CN108763729B (en) | Process industry electromechanical system coupling state evaluation method based on network structure entropy | |
| US20200387818A1 (en) | Asset Optimization Using Integrated Modeling, Optimization, and Artificial Intelligence | |
| US20190147670A1 (en) | Real Time Streaming Analytics for Flight Data Processing | |
| EP2045675B1 (en) | Dynamic management of a process model repository for a process control system | |
| US11409962B2 (en) | System and method for automated insight curation and alerting | |
| CN105467953B (en) | A kind of representation of knowledge and its automation application method towards industrial big data | |
| WO2018204410A1 (en) | Metrology system for machine learning-based manufacturing error predictions | |
| US20020091972A1 (en) | Method for predicting machine or process faults and automated system for implementing same | |
| EP4148523A1 (en) | Intelligent asset anomaly prediction via feature generation | |
| EP3667578A1 (en) | System and method for automatic optimization of a manufacturing bop (bill-of-process) for a production process | |
| CN111650912A (en) | Intelligent manufacturing production management platform for intelligent factory/workshop | |
| CN111340269B (en) | Real-time optimization method for process industrial process | |
| CN116993306A (en) | Knowledge graph-based construction method and device of network collaborative manufacturing system | |
| CN117988988A (en) | Aeroengine fault diagnosis method and system based on feature amplification | |
| CN111832863B (en) | Production scheduling method, system and storage medium for petroleum refining enterprises | |
| CN114789557A (en) | Multi-nozzle printing method of 3d printer | |
| Zhaobing et al. | Application architecture and typical cases of big data technology in health management of civil aircraft system | |
| Crespino et al. | Anomaly detection in aerospace product manufacturing: Initial remarks | |
| CN118311914B (en) | Production line data acquisition control method and system for intelligent workshop | |
| CN118134093B (en) | Intelligent factory quality control system based on machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |