Concurrent analysis driven production line monitoring system

Abstract

The application provides a concurrent analysis driven production line monitoring system, and relates to the technical field of production line monitoring.The system comprises the following steps: equipment operation and control characteristics of an interactive target machine tool production line are determined; hierarchical analysis is used to determine multiple control logic; a penalty function is introduced to determine a micro-processing block and perform node packaging to construct a machine tool monitoring module; a programmable logic controller is introduced to realize automatic configuration and resource management; continuous monitoring is performed to preprocess data at the edge of the equipment and return the data to the monitoring module, concurrent analysis is performed based on a pre-monitoring task node, and production line monitoring is integrated; abnormal sequences are identified, and a man-machine interactive interface is used for visual early warning.The application can solve the technical problem that changes on a production line cannot be responded in time and abnormal events cannot be accurately identified and warned due to insufficient real-time performance and single control logic, and can achieve the technical effect of efficiently and accurately dynamically monitoring a complex and variable machine tool production line and improving the accuracy of identification and early warning of abnormal events.

Description

Technical Field

[0001] This application relates to the field of production line monitoring technology, and in particular to a production line monitoring system driven by concurrent analysis. Background Technology

[0002] With the rapid development of the manufacturing industry and the intensification of market competition, the efficient and stable operation of machine tool production lines is crucial for ensuring product quality, improving production efficiency, and reducing production costs. However, as the scale and complexity of machine tool production lines continue to expand, traditional machine tool production line monitoring methods are no longer adequate to meet the needs of modern manufacturing. Traditional machine tool production line monitoring methods often suffer from insufficient real-time performance and simplistic control logic, resulting in an inability to respond promptly to changes on the production line and difficulty adapting to complex and ever-changing production line environments. Consequently, the identification and early warning of abnormal events are inaccurate, posing potential risks to production safety and quality stability. Summary of the Invention

[0003] The purpose of this application is to provide a concurrent analysis-driven production line monitoring system to solve the technical problems of insufficient real-time performance and simple control logic, which lead to the inability to respond to changes on the production line in a timely manner and make it difficult to accurately identify and warn of abnormal events.

[0004] In view of the above problems, this application provides a production line monitoring system driven by concurrent analysis.

[0005] This application provides a concurrent analysis-driven production line monitoring system, comprising: a production line interaction unit for interacting with the equipment operation and control characteristics of a target machine tool production line, wherein the equipment operation and control characteristics identify machine tool codes; a hierarchical control analysis unit for performing hierarchical control analysis based on the equipment operation and control characteristics to determine multiple control logics, wherein the multiple control logics interact and are associated with machine tool codes; and a node encapsulation unit for introducing penalty functions based on the multiple control logics, performing distributed deployment and supervised training, determining microprocessor blocks and encapsulating nodes to determine the machine tool monitoring module. The system includes: a block; an introduction logic unit for introducing a programmable logic device (PLD) for automating configuration and resource management of the machine tool monitoring module; the PLD being updatable; a concurrent analysis management unit for continuous production line monitoring, preprocessing multi-source production line data at the equipment edge, transmitting it back to the machine tool monitoring module, and performing node concurrent analysis management based on the PLD to integrate and determine the production line monitoring sequence; and an anomaly warning unit for identifying the production line monitoring sequence and determining the anomaly monitoring sequence, providing visualization and anomaly warnings based on a human-machine interface.

[0006] The technical solution provided in this application has at least the following technical effects or advantages:

[0007] The aforementioned concurrent analysis-driven production line monitoring system analyzes the operational characteristics of machine tool production lines to understand equipment operating patterns and demands, and then performs hierarchical control analysis to determine multiple control logics. These control logics ensure appropriate regulation of the production line under different operating conditions. Subsequently, a penalty function is introduced to optimize the control strategy. Through distributed deployment and supervised training, microprocessor blocks are identified and node encapsulated to form independent machine tool monitoring modules. These modules are responsible for monitoring tasks in their respective areas and can respond quickly to problems. Next, a programmable logic device (PLD) is introduced. This PLD can automatically configure and manage resource allocation for the machine tool monitoring modules as needed, and it can also be updated to adapt to production line changes or new monitoring requirements. During production line operation, continuous monitoring is performed, and multi-source production line data is preprocessed at the equipment edge. This preprocessed data is then fed back to the machine tool monitoring modules for further analysis and judgment. Finally, based on the PLD, multiple monitoring tasks are processed concurrently and integrated to determine a single production line monitoring column, providing a comprehensive view of the production line monitoring. Once an abnormal monitoring sequence is identified in the production line monitoring column, an anomaly warning is immediately issued. These warnings are visualized through a human-machine interface. This system achieves comprehensive, efficient, and intelligent monitoring of production line equipment, not only responding to changes in the production line in real time but also accurately identifying and issuing warnings for anomalies, thereby improving production line stability and production efficiency.

[0008] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0010] Figure 1 This is a schematic diagram of the structure of the concurrent analysis-driven production line monitoring system of this application.

[0011] Figure 2 This is a flowchart illustrating the hierarchical control analysis process of the concurrent analysis-driven production line monitoring system described in this application.

[0012] Figure labeling: 1. Production line interaction unit; 2. Hierarchical control and analysis unit; 3. Node encapsulation unit; 4. Logic input unit; 5. Concurrency analysis and management unit; 6. Anomaly warning unit. Detailed Implementation

[0013] This application solves the technical problem of insufficient real-time performance and simple control logic, which leads to an inability to respond to changes on the production line in a timely manner and to accurately identify and warn of abnormal events. It provides a production line monitoring system driven by concurrent analysis, thereby achieving efficient and accurate dynamic monitoring of complex and ever-changing machine tool production lines and improving the accuracy of abnormal event identification and early warning.

[0014] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0015] For examples, please refer to the appendix. Figure 1 This application provides a production line monitoring system driven by concurrent analysis, the system specifically including the following modules:

[0016] Production line interaction unit 1 is used to interact with the equipment operation and control characteristics of the target machine tool production line, wherein the equipment operation and control characteristics identify the machine tool code.

[0017] In this embodiment, the system terminal interacts with the target machine tool production line to understand and identify the operational control characteristics of each piece of equipment. These operational control characteristics refer to the various states, parameters, and performance characteristics exhibited by the equipment during operation. Each of these characteristics has a machine tool code, which corresponds one-to-one with the machine tool equipment. This code is used to distinguish the operational control characteristics of different machine tool equipment, providing basic data for subsequent control and monitoring.

[0018] The hierarchical control analysis unit 2 is used to perform hierarchical control analysis based on the equipment operation and control characteristics to determine multiple control logics, wherein the multiple control logics interact and are associated with machine tool codes.

[0019] In one embodiment, after acquiring the operational characteristics of the machine tool equipment, the system terminal performs hierarchical control analysis. This process involves decomposing complex control requirements into multiple layers based on different dimensions and operating modes, according to the equipment's characteristics and operational requirements, and determining different control logics at each layer. These multiple control logics are connected through inter-layer and intra-layer correlations, enabling different control logics to coordinate with each other based on actual operating conditions, ensuring the smooth operation of the entire production line. Simultaneously, these control logics are also associated with machine tool coding, ensuring that the system terminal can accurately perform personalized control for each machine tool.

[0020] Furthermore, such as Figure 2 As shown, this application provides a method for performing hierarchical control analysis to identify multiple control logics, including:

[0021] Based on static and dynamic dimensions, a first logical layer is determined, and a first-layer logical tuple is mined by combining the equipment operation and control characteristics. Based on independent control and collaborative control, a second logical layer is determined, and a second-layer logical tuple is mined by combining the equipment operation and control characteristics. Collaboration includes synchronous collaboration and sequential collaboration, and the number of logical layers can be expanded. Based on inter-layer correlation and intra-layer correlation, the first-layer logical tuple and the second-layer logical tuple are associated and distributed to determine the multiple control logic.

[0022] Preferably, when constructing the control logic for a machine tool production line, the system terminal determines the first logic layer from both static and dynamic dimensions. The static dimension focuses on the fixed attributes and basic configuration of the equipment, while the dynamic dimension focuses on the operating status and real-time data of the equipment. In determining the first logic layer, the system terminal first collects basic attribute information of the machine tool equipment, such as model, specifications, manufacturer, and design parameters. This information consists of inherent attributes of the equipment and does not change with the operating status. Then, static features related to motion control are extracted from the collected equipment attributes. These features include the equipment's physical characteristics, performance indicators, and design limitations. Subsequently, during machine tool operation, real-time operating status data, such as temperature, pressure, speed, acceleration, current, and voltage, are collected. The collected real-time data is then analyzed and processed to extract dynamic features related to motion control. These features reflect the real-time status, trends, and potential problems of the equipment during operation. Finally, the system terminal merges the static and dynamic features to determine the first logic layer. After determining the first logic layer, the system terminal identifies the main problems that the first logic layer needs to solve, such as basic operation control and safety protection, based on production line requirements and equipment operation and control characteristics. Then, based on the logic requirement analysis and combined with the equipment operation and control characteristics, control logic rules for the first logic layer are formulated. These rules define how the equipment should respond to different inputs and conditions in its basic operating state. Next, the system terminal combines and associates these logic rules to form a layer of logic tuples. This layer of logic tuples refers to data structure units used to characterize the internal relationships and response mechanisms of the control logic layer. Each tuple is composed of feature vectors, logic rules, trigger conditions, response actions, and feedback variables. Feature vectors describe the static and dynamic attribute parameters of the equipment; logic rules define the judgment relationships of the control decision logic; trigger conditions determine the corresponding logic to be executed in which operating state; response actions refer to the specific control instructions executed by the system; and feedback variables record the status information after execution to achieve closed-loop control. A single-layer logical tuple can be stored using a table structure or a key-value mapping structure, and can be associated in the system as graph nodes or matrices. It supports calculation and reasoning, and is used to achieve specific control objectives, representing a preliminary control strategy based on the basic attributes and operating status of the device.

[0023] After obtaining the first-level logical tuple, the system terminal begins to determine the second logical layer, which focuses on independent control and collaborative control. Independent control optimizes a single device or process step, while collaborative control considers the interaction between multiple devices or process steps. In collaborative control, the system terminal further subdivides into synchronous collaboration (multiple devices or steps working simultaneously) and sequential collaboration (devices or steps working in a specific order). In determining the second logical layer, the system terminal analyzes which devices or steps require independent control and which require collaborative control based on the results of the first logical layer. For example, when a machine tool needs to be stopped urgently for safety reasons, independent control is used to ensure rapid response and reduce potential losses. Subsequently, based on the analysis results, the system terminal determines the scope, relationship, and interaction mode of independent and collaborative control, thus determining the second logical layer. Next, key operational control features related to the second logical layer are extracted from the device's operational control characteristics. These features involve the device's performance, status, response speed, etc. Based on the extracted key operational control features, logical rules for independent and collaborative control are formulated. For collaborative control, the system terminal also considers rules for synchronous and sequential collaboration. Then, the system terminal combines and associates the defined logical rules to form a second-level logical tuple. This second-level logical tuple is similar to the first-level logical tuple, each containing a set of related logical rules and conditions to achieve specific control objectives. However, these tuples are more complex, taking into account the interdependencies and collaboration between devices. In designing the second logical layer, the system terminal also reserves expansion interfaces to allow for the addition of new logical layers as needed.

[0024] After determining the first and second logical layers, the system terminal analyzes the logical coherence between the first-layer and second-layer logical tuples. This includes determining their dependencies, influencing factors, and the direction of information flow, and assessing the functional complementarity of the two logical layers. For example, the first-layer logical tuple might focus on basic control and stability, while the second-layer logical tuple might emphasize optimization and synergy. Subsequently, within each logical layer, the logical relationships between tuples are analyzed, including inter-layer correlation analysis and intra-layer correlation analysis. For inter-layer correlation analysis, a parameter mapping matrix is ​​established to match the output variables of the first-layer logical tuple with the input parameters of the second-layer logical tuple, and their Pearson correlation coefficient is calculated. If the Pearson correlation coefficient exceeds a preset threshold, such as 0.7, strong inter-layer correlation is determined. For intra-layer correlation analysis, feature weight analysis and logical coupling degree calculation methods are used to quantify the shared feature parameters, control objectives, and constraints between each tuple. For example, by calculating the feature overlap or rule co-occurrence frequency between tuples, an intra-layer correlation matrix is ​​obtained, thereby distinguishing between independent tuples and strongly coupled tuples. Subsequently, based on the analysis of inter-layer and intra-layer correlations, an association distribution operation is performed. That is, using a graph model or tensor distribution algorithm, each logical tuple is treated as a node, and inter-layer or intra-layer correlations are treated as weighted connection edges. The distribution location and connection strength of the logical nodes are determined by minimizing global coupling error or maximizing information flow efficiency, thus forming an association network between layers one and two. The intra-layer relationships in this association network reflect the coupling structure of local control logic, while the inter-layer relationships embody the logical transmission paths between different layers, clearly reflecting the information flow and control relationships between two logical layers. Then, the system terminal integrates the inter-layer association network and intra-layer logical relationships to form a complete multi-layer control logic. This multi-layer control logic can cover all aspects from basic control to collaborative optimization and can be dynamically adjusted according to the real-time status of the equipment and production line requirements to achieve efficient and stable production line operation.

[0025] Node encapsulation unit 3 is used to introduce a penalty function based on the multiple control logic, perform distributed deployment and supervised training, determine the microprocessor block and encapsulate the node, and determine the machine tool monitoring module.

[0026] In one embodiment, to ensure monitoring stability, a penalty function is introduced in the system terminal. This penalty function is a mechanism for measuring system performance deviations, constructed based on the localization of interference points in the interpretation of multiple control logics. When deviations occur during actual operation, the penalty function applies a corresponding penalty to guide the system terminal in correcting biases in data analysis. Subsequently, the system terminal distributes the associated distribution of multiple control logics in a distributed manner. Distributed deployment means distributing the various control logics of the system terminal across different physical locations to achieve better load balancing and scalability. By distributing the multiple control logics, the system terminal determines the distribution of microprocessor blocks. These microprocessor blocks refer to the smallest computing units partitioned for performing data processing tasks, i.e., independently runnable computing logic nodes or lightweight task containers, specifically lightweight task instances on embedded processing cores, virtual computing threads, or programmable logic units. Then, based on the distributed deployment, the system terminal uses sample data to perform supervised training on the microprocessor block distribution, enabling it to learn and adapt to various working conditions and scenarios. Through continuous iteration and optimization, its monitoring and control capabilities for machine tool equipment can be gradually improved. As training progresses, the system terminal identifies microprocessor blocks that meet convergence criteria and encapsulates them into nodes. This node encapsulation reduces the complexity and dependencies of the system terminal, improving its maintainability and scalability. The system terminal then connects these encapsulated nodes to form the machine tool monitoring module. This module is the core of the entire system terminal, responsible for receiving data and information from various nodes and performing comprehensive analysis and processing. Guided by multiple control logics and penalty functions, the machine tool monitoring module can monitor the machine tool's operating status in real time and predict potential problems to ensure stable operation and efficient production.

[0027] Furthermore, this application provides the aforementioned penalty function for performing data interpretation mainline bias backreference, including:

[0028] The data interpretation main line is determined by traversing the multiple control logics, and the data interpretation main line corresponds one-to-one with the multiple control logics; based on the data interpretation main line, the interpretation interference points based on multi-source production line data are located; for the interpretation interference points, the penalty function is constructed and triggered for application management.

[0029] Preferably, to ensure the accuracy of data interpretation and avoid biased data guidance, the system terminal performs logical analysis to assess the monitoring status. In this process, the system terminal first traverses multiple control logics to identify key nodes that directly affect the system's state, behavior, or performance. These nodes are turning points or decision points in the logic chain. Subsequently, the data sources and data types that each key control node depends on are analyzed. This data is necessary for executing the control logic and forms the basis of data interpretation. Then, based on the control logic and data dependencies, data interpretation paths are constructed from the data source to the control output. These paths describe how data flows within the system terminal and how it is processed, analyzed, and interpreted at different control nodes. Finally, among all the constructed data interpretation paths, the main data interpretation paths, i.e., the main data interpretation lines, are identified. These main data interpretation lines refer to the dominant data flow paths formed between the original multi-source production line data input and the result output; that is, the core transmission and computation routes of data between multiple control layers and algorithms. They correspond one-to-one with multiple control logics, ensuring that data can be accurately interpreted according to the predetermined logical paths. Subsequently, based on these established data interpretation mainlines, the system terminal identifies potential interpretation interference points during the multi-source production line data interpretation process. These interference points may be caused by factors such as the diversity of data sources, instability in data quality, or logical conflicts. They can cause data interpretation to deviate from the correct mainline, exhibiting a turning point effect. These turning points can be detected and identified through residual analysis, such as when the deviation between the actual output and the predicted output exceeds a set threshold. To address these interpretation interference points, the system terminal constructs a penalty function and manages its application. The form of the penalty function can be customized according to specific circumstances, such as simple weight adjustments, increasing the loss function, or more complex nonlinear mappings. The goal is to correct the data interpretation process when a deviation from the mainline is detected. That is, when a deviation from the mainline is detected, the penalty function feedback adjustment mechanism, which performs negative feedback correction on the weights of the mainline nodes, pulls the calculation results of the data flow back onto the mainline trajectory. This bias correction mechanism can ensure the accuracy and consistency of data interpretation, avoid erroneous results caused by biased guidance of some data, and ensure the accuracy and reliability of data analysis.

[0030] Furthermore, this application provides the aforementioned machine tool monitoring module, comprising:

[0031] Based on the associated distribution of the multiple control logics, a distributed deployment is performed to determine the distribution of microprocessor blocks; industrial big data retrieval is performed to determine and segment sample data, which is then mapped to the distribution of microprocessor blocks for supervised training and cross-validation to determine the microprocessor blocks that meet the convergence conditions; each microprocessor block is encapsulated as a node, and a fast interaction channel is established based on the associated distribution to generate the machine tool monitoring module.

[0032] Preferably, the system terminal divides the control logic into different microprocessor blocks based on the interrelationships between multiple control logics—that is, which control logics are interdependent and which can operate independently—to achieve distributed deployment. Distributed deployment helps improve the system's scalability, flexibility, and fault tolerance. Subsequently, industrial big data retrieval technology is used to extract sample data related to machine tool monitoring from massive amounts of data. This sample data is divided into multiple parts for subsequent training and validation of the microprocessor block distribution. Then, the system terminal maps the segmented sample data to the determined microprocessor block distribution and performs supervised training on each microprocessor block. Specifically, taking a random microprocessor block as an example, the system terminal sets corresponding hyperparameters for this microprocessor block, such as learning rate, number of iterations, batch size, etc. Then, the mapped sample data is divided into k equal-sized subsets, where k is determined by comprehensively considering the evaluation's stability, accuracy, and computational cost; commonly used k values ​​include 5 and 10. Next, the system terminal randomly selects one of the k subsets as the validation set, and the rest as the training set. The training set is then used to train the corresponding microprocessor block. Each microprocessor block learns the mapping relationship from input data to output results. By iteratively optimizing the parameters of the microprocessor block, its prediction results gradually approach the true results. Then, the trained microprocessor block is validated using the validation set, and its accuracy is calculated. This process is repeated until each subset has been used as a validation set. After training, the system terminal extracts the accuracy from the k training iterations, compares them, and selects the training result with the highest accuracy. It then compares this result with a preset convergence condition. If the convergence condition is met, the system terminal uses this training result as the final microprocessor block. If the convergence condition is not met, the system terminal re-divides the subsets and repeats the training and validation process until the convergence condition is satisfied.

[0033] After training the microprocessor blocks, the system terminal encapsulates each microprocessor block that meets the convergence criteria as an independent node for easy management and integration. Simultaneously, based on the distributed relationships of multiple control logics, the dependencies and communication requirements between the microprocessor blocks are determined. Subsequently, a suitable communication protocol is selected to establish a fast interactive channel according to actual needs. For example, if real-time performance is required, the system terminal chooses UDP for establishing the fast interactive channel because UDP is a connectionless protocol that does not establish a connection before transmitting data, thus offering lower latency. Afterward, the system terminal integrates the encapsulated microprocessor blocks and the fast interactive channel to generate a machine tool monitoring module. This module can monitor the machine tool's operating status in real time and predict potential faults. Through distributed deployment and the collaborative work of the microprocessor blocks, the machine tool monitoring module can achieve efficient and accurate data processing and analysis, improving the reliability and production efficiency of the machine tool.

[0034] Furthermore, this application provides the aforementioned convergence conditions, including:

[0035] The convergence condition is that the verification samples mapped by the microprocessor block satisfy the upper limit of verification accuracy, and the verification samples of any block not mapped by the microprocessor block satisfy the lower limit of verification accuracy.

[0036] Optionally, to more accurately evaluate the trained microprocessor blocks, the system terminal pre-sets a convergence condition. The upper limit of this convergence condition is determined by the mapping of validation samples to the required accuracy, while the lower limit is determined by the requirement that validation samples from any non-mapped block meet the required accuracy. This convergence condition ensures that the validation accuracy of samples from the block is sufficiently high, while the validation accuracy of samples from other blocks is sufficiently low, improving the block's specificity and ensuring differentiation between blocks. When a trained microprocessor block satisfies this convergence interval, it means that the data processing and analysis within that block accurately reflects the actual situation, thus guaranteeing the effectiveness and accuracy of the block's function.

[0037] A logic unit 4 is introduced to introduce a programmable logic device (PLD). The PLD is used to perform automated configuration and resource configuration management of the machine tool monitoring module. The PLD is updatable.

[0038] In one embodiment, the system terminal introduces and configures a programmable logic controller (PLC), then uses the configured PLC as a management tool for the machine tool monitoring module. This aims to achieve automated configuration and resource allocation management guided by monitoring dimensions and production line processing tasks. The PLC's updatable nature enables efficient and flexible global coordination and automated control of machine tool monitoring across the entire system. This tool allows the system terminal to more easily update the configuration, allocate resources, and optimize the performance of the machine tool monitoring module to meet the needs of different production line processing tasks, thereby improving the automation level and efficiency of the entire production line.

[0039] Furthermore, this application provides the aforementioned introduced programmable logic device, including:

[0040] The monitoring standards of the target machine tool production line based on work order production tasks are read; based on the architecture of the machine tool monitoring module, the logical flow based on the monitoring standards is determined, and a periodic control program is set, wherein the period is the module processing cycle of single-phase monitoring and the time zone monitoring cycle of work order product processing; based on the periodic control program, the programmable logic device is configured.

[0041] Preferably, to effectively monitor the production tasks of the target machine tool production line, the system terminal first reads the monitoring standards upon which these tasks are based. These standards include specific requirements for production quality, efficiency, and safety. Subsequently, based on the existing architecture of the machine tool monitoring module, a logical flow is determined. This flow follows the previously read monitoring standards and covers the entire monitoring process from data acquisition and processing to result output and response. Next, to ensure the real-time performance and accuracy of monitoring, the system terminal sets a periodic control program. This period includes two aspects: first, the module processing cycle for single-phase monitoring, representing the time interval at which each monitoring module collects and processes data, set based on the frequency of data updates and the complexity of processing; second, the time zone monitoring cycle related to the processing of the work order products, set based on the actual processing flow and time requirements of the products. With this periodic control program in place, the system terminal configures the programmable logic controller (PLC). Through configuration, the PLC can automatically monitor and manage the machine tool production line according to the predetermined logical flow and periodic control program, ensuring the smooth progress of production tasks. This process combines monitoring standards with the architecture of machine tool monitoring modules. By setting periodic control programs, it enables the configuration of programmable logic devices to achieve automated monitoring and management of production tasks for target machine tool production lines.

[0042] Concurrent analysis management unit 5 is used to continuously monitor the production line, preprocess multi-source production line data at the equipment edge, and send it back to the machine tool monitoring module. Based on the programmable logic device, it performs node concurrent analysis management based on pre-monitoring tasks and integrates and determines the production line monitoring column.

[0043] In one embodiment, during continuous production line monitoring, the system terminal performs preprocessing of multi-source production line data at the equipment edge. That is, before the data is transmitted to the central machine tool monitoring module, the edge device performs preliminary processing and filtering of data from different sources to ensure data accuracy and validity. Subsequently, the edge device transmits the processed data to the machine tool monitoring module via a set network communication protocol. Afterwards, the system terminal, based on a programmable logic device (PLC), performs concurrent node analysis and management based on the pre-monitoring task. Specifically, the system terminal uses the PLC to segment and activate nodes of the multi-source production line data, and performs concurrent processing and result combination of multiple nodes, thereby quickly responding to various situations on the production line and determining the production line monitoring column. The production line monitoring column provides a comprehensive display of the overall operation of the production line, including real-time data and status information of each key monitoring node, enabling the system terminal to understand the production line's operating status and make corresponding decisions.

[0044] Furthermore, this application provides an interface between the machine tool monitoring module at the device edge and the backend based on a network communication protocol, including:

[0045] The multi-source production line data is traversed, and based on the preprocessing model, data isomorphism processing and multi-source feature fitting are performed to determine valid production line data. The isomorphic data structure can be any of the multi-source data structures. The valid production line data is then fed back, and node activation and concurrent analysis management are performed based on the programmable logic device to determine the production line monitoring column. The production line monitoring column is identified, and the column coefficients are binarized based on the mapping monitoring threshold. Inferior binary data is identified and marked, and the monitoring threshold corresponds one-to-one with the production line monitoring column.

[0046] Preferably, during continuous production line monitoring, the system terminal traverses multi-source production line data from various types of sensors and monitoring equipment. To fully utilize this data and avoid data redundancy, the system terminal performs isomorphic processing and multi-source feature fitting on this multi-source data based on a preprocessing model. This preprocessing model is customized according to the specific needs, data types, and purposes of the production line, integrating necessary preprocessing algorithms. Isomorphic processing transforms data from different sources into a unified data structure. This involves selecting one of these multi-source data structures as a baseline and transforming the remaining data to ensure consistency and comparability. Through multi-source feature fitting, the system terminal can integrate and optimize data from different sources for the same objective, thereby filtering out effective production line data. This process not only improves data transmission efficiency but also ensures data accuracy and reliability. Subsequently, the system terminal sends this effective production line data back to the machine tool monitoring module. Here, based on a programmable logic device (PLC), the system terminal performs node activation and concurrent analysis management on this data, simultaneously processing multiple monitoring nodes and responding and analyzing rapidly according to set logical rules. In this way, the system terminal can monitor the production line's operation in real time and identify individual monitoring columns, which provide a comprehensive display of key monitoring points across the production line. Then, for each monitoring column, parameters such as temperature, pressure, current, and speed are normalized from maximum to minimum to eliminate dimensional differences. A weighted sum is then performed according to preset weights to obtain the column coefficient for each monitoring column. A higher column coefficient indicates a greater deviation from the expected operating parameters. Once the column coefficient is determined, the system terminal performs binarization based on mapped monitoring thresholds. This involves using statistical analysis to determine the normal fluctuation range as the baseline threshold (e.g., the mean plus or minus one standard deviation). The calculated column coefficient is then compared to the corresponding baseline threshold. When the column coefficient falls within the baseline threshold, a logic value of 1 is output, indicating a normal state; when the coefficient exceeds the baseline threshold, a logic value of 0 is output, indicating an abnormal state. In this way, the system terminal can quickly identify abnormal situations on the production line and take appropriate measures to ensure stable operation.

[0047] Furthermore, this application provides node activation and concurrency analysis management based on the programmable logic device, including:

[0048] Based on the pre-monitoring task, the effective production line data is segmented and nodes are activated. Multi-node concurrent processing and result combination are performed to determine the production line monitoring column. The concurrent processing includes multi-task concurrent processing and single-task multi-node concurrent processing. If it is multi-task concurrent processing, the node processing domain exceeds the limit based on the task time limit. If the limit is exceeded, the microprocessor block is temporarily reallocated based on the operator utilization rate.

[0049] Optionally, in production line monitoring, after the system terminal obtains pre-processed and valid production line data, the next step is to process this data according to the pre-monitoring tasks. This process includes data segmentation, node activation, and concurrent processing by multiple nodes. First, the system terminal segments this valid production line data according to the requirements of the pre-monitoring tasks, that is, assigns the data to different processing units or nodes. Then, these nodes are activated, enabling them to begin processing their respective data. To improve processing efficiency, the system terminal adopts a multi-node concurrent processing approach. This means that multiple nodes can process data simultaneously, thereby greatly accelerating the processing speed. This concurrent processing can be multi-task concurrent processing, i.e., processing multiple different tasks simultaneously; or single-task multi-node concurrent processing, i.e., a task is divided into multiple sub-tasks, which are processed simultaneously by different nodes. After each node completes data processing, the system terminal combines the processing results of these nodes to form a complete production line monitoring column. This monitoring column provides a comprehensive reflection of the entire production line's operational status. In multi-task concurrent processing, since each task has its specific processing time limit, the system terminal monitors the processing progress of the nodes. If a node's processing time exceeds its task time limit, indicating a processing domain overload, the system terminal temporarily reallocates resources based on operator utilization. This operator utilization is the ratio of the number of operator calls a node makes during task execution to its maximum supported number of calls, used to measure the node's real-time computing load. Based on this operator utilization, the system terminal can temporarily allocate some resources from other relatively idle nodes to the overloaded node. Allocation rules can include prioritizing idle or lightly loaded nodes with current operator utilization below a set threshold as resource providers; selecting the node with the lowest cost for collaborative load sharing based on data channel bandwidth and topology location between nodes; and setting a resource reallocation frequency cap, allowing only one resource reallocation within the same task cycle, thereby helping the node complete its task as quickly as possible and ensuring the stability and processing efficiency of the entire system.

[0050] The anomaly warning unit 6 is used to identify the production line monitoring column and determine the anomaly monitoring sequence, and to perform visualization and anomaly warning based on the human-machine interface.

[0051] In one embodiment, during production line monitoring, after effective production line data undergoes segmentation, node activation, and multi-node concurrent processing, one or more production line monitoring columns are obtained. These columns specifically reflect the production line's operational status, containing a large amount of data and information. Subsequently, the system terminal identifies potential abnormal monitoring sequences from these monitoring columns and visualizes and issues warnings through a human-machine interface. To identify abnormal monitoring sequences, the system terminal compares each production line monitoring column with a standard production line monitoring column, identifying data points or segments that do not conform to the normal operating mode, thereby determining whether an anomaly exists. Once an abnormal monitoring sequence is identified, the system terminal presents this abnormal information visually through the human-machine interface. Visualization includes various forms such as charts and curves, aiming to intuitively display the production line's operational status and abnormal situations. Through visualization, the overall situation of the production line can be quickly understood, and potential problems can be identified and addressed promptly. When the system terminal detects an abnormal monitoring sequence, it also automatically triggers an alert function, sending alert information to the user via sound, light, email, or SMS. In this way, the user can be informed of the abnormal situation immediately and take appropriate measures to handle it. Anomaly warnings can help users identify and address potential risks in a timely manner, ensuring the stable operation of the production line.

[0052] In summary, the concurrent analysis-driven production line monitoring system provided in this application has the following technical advantages:

[0053] This application collects the equipment operation and control characteristics of the target machine tool production line and determines multiple control logics based on these characteristics. Subsequently, a penalty function is introduced to perform distributed deployment and supervised training of the multiple control logics, forming microprocessor blocks and encapsulating them into nodes to construct a machine tool monitoring module. Next, a programmable logic device (PLD) is introduced to enable automated configuration and management of the machine tool monitoring module to adapt to the changing needs of the production line. Then, multi-source production line data is preprocessed at the equipment edge and managed through concurrent node analysis using the PLD, integrating and determining a single monitoring column for the production line. Finally, by identifying abnormal sequences in the monitoring column and providing visualization and early warning, users can promptly discover and address potential problems. These technical effects collectively solve the technical problems of insufficient real-time performance and single control logic, which lead to an inability to respond promptly to changes on the production line and difficulty in accurately identifying and warning of abnormal events. This achieves efficient and accurate dynamic monitoring of complex and ever-changing machine tool production lines, improving the accuracy of abnormal event identification and early warning.

[0054] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0055] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A production line monitoring system driven by concurrency analysis, characterized in that, The system includes: The production line interaction unit is used to interact with the equipment operation and control characteristics of the target machine tool production line, wherein the equipment operation and control characteristics identify the machine tool code; The hierarchical control analysis unit is used to perform hierarchical control analysis based on the equipment operation and control characteristics to determine multiple control logics, wherein the multiple control logics interact and are associated with machine tool codes. The node encapsulation unit is used to introduce a penalty function based on the multiple control logic, perform distributed deployment and supervised training, determine the microprocessor block and encapsulate the node, and determine the machine tool monitoring module. A logic unit is introduced to introduce a programmable logic device (PLD). The PLD is used to perform automated configuration and resource configuration management of the machine tool monitoring module. The PLD is updatable. The concurrent analysis and management unit is used to continuously monitor the production line, preprocess multi-source production line data at the equipment edge, and send it back to the machine tool monitoring module. Based on the programmable logic device, it performs node concurrent analysis and management based on the pre-monitoring task and integrates and determines the production line monitoring column. An anomaly warning unit is used to identify the production line monitoring column and determine the anomaly monitoring sequence, and to perform visualization and anomaly warning based on the human-machine interface. Identify multiple control logics, including: Based on static and dynamic dimensions, the first logical layer is determined, and a logical tuple is mined by combining the device operation and control characteristics. Based on independent control and collaborative control, a second logical layer is determined, and the second-layer logical tuples are mined by combining the equipment operation and control characteristics. Among them, collaboration includes synchronous collaboration and sequential collaboration, and the number of logical layers can be expanded. Based on inter-layer correlation and intra-layer correlation, the first-layer logic tuple and the second-layer logic tuple are associated and distributed to determine the multiple control logic; The introduced penalty function is used for data interpretation mainline bias back-reference, including: By traversing the multiple control logics, the main data interpretation line is determined, and the main data interpretation line corresponds one-to-one with the multiple control logics. Based on the main data interpretation line, the interpretation interference points based on multi-source production line data are located; For the interpretation interference points, the penalty function is constructed and triggered for application management.

2. The concurrent analysis-driven production line monitoring system as described in claim 1, characterized in that, The machine tool monitoring module includes: Based on the associated distribution of the multiple control logics, a distributed deployment is carried out to determine the distribution of microprocessor blocks; Industrial big data retrieval is performed to identify and segment sample data, which is then mapped to the distribution of the microprocessor blocks for supervised training and cross-validation to determine the microprocessor blocks that meet the convergence conditions. The machine tool monitoring module is generated by encapsulating single microprocessor blocks as nodes and establishing a fast interaction channel based on the associated distribution.

3. The concurrent analysis-driven production line monitoring system as described in claim 2, characterized in that, The convergence condition is that the verification samples mapped by the microprocessor block satisfy the upper limit of verification accuracy, and the verification samples of any block not mapped by the microprocessor block satisfy the lower limit of verification accuracy.

4. The concurrent analysis-driven production line monitoring system as described in claim 1, characterized in that, The introduction of the programmable logic device includes: Read the monitoring standards of the target machine tool production line based on work order production tasks; Based on the architecture of the machine tool monitoring module, the logical flow based on the monitoring standard is determined, and a periodic control program is set, wherein the period is the module processing cycle of single-phase monitoring and the time zone monitoring cycle of work order product processing. The programmable logic device is configured based on the cycle control program.

5. The concurrent analysis-driven production line monitoring system as described in claim 1, characterized in that, The machine tool monitoring module at the device edge and the back end interact based on a network communication protocol, including: The multi-source production line data is traversed, and based on the preprocessing model, data isomorphism processing and multi-source feature fitting are performed to determine the effective production line data. The isomorphic data structure can be any one of the multi-source data structures. The valid production line data is returned, and node activation and concurrent analysis management are performed based on the programmable logic device to determine the production line monitoring column; Identify the production line monitoring column, perform binarization processing on the column coefficients based on the mapped monitoring threshold, identify and mark the inferior binary data, and the monitoring threshold corresponds one-to-one with the production line monitoring column.

6. The concurrent analysis-driven production line monitoring system as described in claim 5, characterized in that, Node activation and concurrency analysis management based on the programmable logic device include: Based on the pre-monitoring task, the effective production line data is segmented and nodes are activated, multi-node concurrent processing and result combination are performed, and the production line monitoring column is determined. The concurrent processing includes multi-task concurrent processing and single-task multi-node concurrent processing. If multiple tasks are running concurrently, the node processing domain exceeds the limit based on the task time limit; If the limit is exceeded, the resources of the microprocessor block will be temporarily reallocated based on the operator utilization rate.

Images