Abnormity detection and processing method for production process of intelligent factory

Abstract

The invention belongs to the technical field of lean manufacturing monitoring, and particularly relates to an anomaly detection and processing method for a production process of an intelligent factory, and the method comprises the steps: constructing a dual monitoring node chain and a layered anomaly association evaluation node chain based on a preset production process flow chain, and collecting the operation state and process parameter data of equipment in real time; the hierarchical anomaly association assessment node chain analyzes the equipment health degree through the operation state assessment sub-chain, mines a cross-process anomaly propagation relationship through the whole-process technology anomaly assessment sub-chain, generates a whole-process association anomaly assessment space, fuses real-time data and historical anomaly strategy index information based on a distributed algorithm, and finally performs the whole-process association anomaly assessment on the basis of the real-time data and the historical anomaly strategy index information. Dynamically constructing a difference exception processing strategy library, and realizing processing strategy synchronous feedback of positioning exception; according to the invention, through a dual monitoring architecture and hierarchical association evaluation analysis, the problems of fuzzy abnormal propagation path and response lag are solved, and the real-time performance and processing accuracy of abnormal detection are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of lean manufacturing monitoring, and in particular relates to a method for detecting and processing anomalies in a production process of an intelligent factory. Background Art

[0002] As high-end manufacturing industries place increasing demands on the quality of high-precision steel pipes and fittings, lean manufacturing-based production processes urgently need to implement refined management and control over the entire life cycle. The production of seamless pipe fittings involves a complex process encompassing more than ten steps, including blanking, pickling, and solutionizing. Even slight deviations in process parameters (such as acid concentration fluctuations or dimensional tolerance violations) in each step can trigger cascading anomalies across multiple steps, leading to batch scrapping or production line paralysis. However, existing anomaly detection technologies suffer from systemic flaws: First, isolated detection and delayed response rely on single-point sensors or manual spot checks, lacking inter-process data collaboration. For example, an anomaly in the pickling process is only identified during the subsequent shot blasting or finished product inspection phase, resulting in a delay in timely detection and intervention. Second, localization is difficult, as tracing the root cause of anomalies relies on manual judgment and experience, making it impossible to quantify the specific impact of anomalies in previous steps on subsequent steps (for example, how deviations in forming process parameters lead to shot blasting defects), resulting in inefficient troubleshooting. Third, static thresholds are limited, and fixed alarm rules cannot adapt to dynamic production conditions (such as raw material batch differences and equipment performance degradation). This leads to frequent false alarms or missed detections, impacting production continuity. To address this issue, the present invention provides a method for detecting and handling anomalies in smart factory production processes. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention proposes a method for anomaly detection and processing in the production process of an intelligent factory, the method comprising: constructing a dual monitoring node chain and a hierarchical anomaly association evaluation node chain based on a preset production process flow chain, and collecting equipment operating status and process parameter data in real time; the hierarchical anomaly association evaluation node chain analyzes the equipment health through the operating status evaluation sub-chain, and mines the cross-process anomaly propagation relationship through the full-process process anomaly evaluation sub-chain to generate a full-process associated anomaly evaluation space, and at the same time, based on a distributed algorithm to fuse real-time data and historical anomaly strategy index information, dynamically construct a differential anomaly processing strategy library to achieve synchronous feedback of the processing strategy for located anomalies; this application breaks through the limitations of traditional isolated detection through a dual monitoring architecture and hierarchical association analysis, solves the problems of fuzzy anomaly propagation paths and delayed responses, and improves the real-time performance and processing accuracy of anomaly detection.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] A method for detecting and processing anomalies in a smart factory production process, comprising:

[0006] Based on the preset production process chain, a dual monitoring node chain and a hierarchical abnormality correlation evaluation node chain are constructed;

[0007] The hierarchical abnormality association evaluation node chain includes an operation status evaluation subchain and a full-process process abnormality evaluation subchain;

[0008] Based on the dual monitoring node chain, the production process chain is monitored in real time to obtain equipment operation status information and production process information of the equipment to be processed;

[0009] Based on the equipment operation status information and the production process information of the equipment to be processed, combined with the hierarchical abnormality association evaluation node chain, the whole process associated abnormality evaluation space is obtained;

[0010] Based on the full-process associated anomaly evaluation space, an anomaly handling strategy library is constructed through a distributed algorithm combined with historical anomaly strategy index information to obtain a differential anomaly handling strategy, and the differential anomaly handling strategy is fed back to the dual monitoring node chain in real time to provide real-time warning and processing for the anomalies monitored and located in the production process chain.

[0011] Specifically, the production process chain includes a process node sub-chain and an equipment operation node sub-chain; the dual monitoring node chain includes an equipment operation monitoring sub-chain and a process monitoring sub-chain;

[0012] Each processing node in the process node subchain stores the processing parameters of the corresponding node; each operation node in the equipment operation node subchain stores the equipment control parameters for the corresponding processing equipment to execute the processing technology;

[0013] The equipment operation monitoring subchain has a one-to-one correspondence with the nodes in the equipment operation node subchain; the process node subchain has a one-to-one correspondence with the nodes in the process monitoring subchain;

[0014] Each monitoring node in the equipment operation monitoring subchain corresponds one-to-one to an operation evaluation node in the operation status evaluation subchain.

[0015] Specifically, each process monitoring node in the process monitoring subchain corresponds one-to-one to each process assessment node in the full-process process anomaly assessment subchain;

[0016] Each operation evaluation node in the operation status evaluation subchain is connected to the process evaluation node in the full process process abnormality evaluation subchain in a one-to-many backward mapping manner;

[0017] Each operation evaluation node is configured with a hierarchical operation evaluation index set for evaluating the operation status of the corresponding processing equipment; each process evaluation node is configured with a hierarchical process evaluation index set for evaluating the corresponding processing technology;

[0018] The operation hierarchical evaluation index set and the hierarchical process evaluation index set are respectively stored in a preset hierarchical operation index library and a hierarchical process index library, and a one-to-many directed connection is performed between the hierarchical process index library and the hierarchical operation index library through a preset directed association connection.

[0019] Specifically, the construction process of the production process chain includes:

[0020] Obtain the full-process production process information of each device to be processed and combine it with the pre-trained production process analysis algorithm to obtain the continuous processing process node sequence and processing order information corresponding to each device to be processed, as well as the required processing equipment sequence and processing equipment operation interaction information, as well as the mapping information and associated causal information of each processing technology and processing equipment;

[0021] Based on the continuous processing node sequence and processing order information corresponding to each device to be processed, a loosely coupled algorithm is used to construct a directed loosely coupled connection between each processing node corresponding to each device to be processed;

[0022] Based on the continuous processing process node sequence and the directed loosely coupled connection, obtaining a process node sub-chain corresponding to each device to be processed;

[0023] At the same time, the equipment operation node sequence is constructed based on the required processing equipment sequence, and the interactive information of each processing equipment operation is used to build a two-way operation abnormality reasoning connection through the graph neural network;

[0024] Based on the device operation node sequence combined with the bidirectional operation abnormality reasoning connection, a device operation node sub-chain is obtained.

[0025] Specifically, the construction process of the production process chain also includes:

[0026] Assume that the current process node sub-chain contains M continuous processing nodes, the equipment operation node sub-chain contains N equipment operation nodes, and the process parameters corresponding to each processing node are only processed and run on one equipment operation node;

[0027] Based on the assumption, the processing node corresponding to the process parameters running on each equipment operation node is set as the main processing node of the equipment operation node, and the remaining continuous processing nodes after the main processing node constitute the causal relationship process node sequence of the current equipment operation node;

[0028] Based on the main causal correlation information between the operating status and processing results of each equipment operation node and the corresponding main processing node, the main correlation degree between each equipment operation node and the corresponding main processing node is constructed through the autocorrelation algorithm, and the main connection relationship between the corresponding equipment operation node and the main processing node is constructed using the obtained main correlation degree;

[0029] At the same time, based on the collaborative operation causal information between each equipment operation node and the corresponding causal process node sequence, the collaborative causal correlation degree of each equipment operation node for each processing node in the corresponding causal process node sequence is obtained through the cross-correlation algorithm.

[0030] Specifically, the construction process of the production process chain also includes:

[0031] Based on the collaborative causal correlation degree of each equipment operation node to each processing node of the corresponding causal correlation process node sequence, a collaborative causal process connection is constructed between the equipment operation node and the corresponding causal correlation process node sequence;

[0032] Connect the nodes corresponding to the process node subchain and the equipment operation node subchain based on the main connection relationship and the collaborative causal process connection to obtain a production process flow chain;

[0033] A causal reasoning model is constructed based on the Bayesian algorithm and configured into the bidirectional operation abnormality reasoning connection and the collaborative causal process connection to obtain a production process flow chain with causal reasoning.

[0034] Specifically, the construction process of the dual monitoring node chain includes:

[0035] Based on the architecture mapping of the device operation node subchain, a device operation monitoring subchain with the same architecture is obtained, and monitoring equipment is configured according to the corresponding operation attributes of each device operation node to obtain the device operation parameter space corresponding to the device operation node subchain;

[0036] The device operation parameter space includes a single operation parameter sequence corresponding to each device operation node and an interactive operation parameter sequence corresponding between each device operation node;

[0037] At the same time, based on the architecture mapping of the process node sub-chain, a process monitoring sub-chain with the same architecture is obtained, and the corresponding monitoring equipment is configured according to the processing technology information to obtain the processing technology monitoring information space corresponding to the equipment operation node sub-chain.

[0038] Specifically, the process of building a hierarchical anomaly association evaluation node chain includes:

[0039] Based on the architecture mapping of the equipment operation node subchain, the operation status evaluation subchain of the same architecture is obtained, and based on the architecture mapping of the process node subchain, the full-process process anomaly evaluation subchain of the same architecture is obtained;

[0040] The main connection relationship and the collaborative causal process connection in the production process flow chain with causal reasoning are mapped to the corresponding nodes in the operation status evaluation subchain and the full-process process anomaly evaluation subchain, and a one-to-many backward mapping connection between the operation status evaluation subchain and the full-process process anomaly evaluation subchain is constructed to obtain a hierarchical anomaly association evaluation node chain.

[0041] Specifically, the process of obtaining differential exception handling strategies includes:

[0042] Based on the dual monitoring node chain, the equipment operation parameter space and the processing technology monitoring information space are monitored in real time, and synchronously mapped to the corresponding operation status assessment subchain and full-process process anomaly assessment subchain in the hierarchical anomaly association assessment node chain;

[0043] Based on the hierarchical operation evaluation indicator set configured for each operation evaluation node in the operation status evaluation subchain and the full-process process abnormality evaluation subchain, and the hierarchical process evaluation indicator set configured for each process evaluation node, the operation status and processing process results of each operation evaluation node and the corresponding main processing process node are synchronously evaluated to obtain an operation status evaluation result sequence and a processing process evaluation result sequence;

[0044] Set the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence. Based on the operating status assessment result sequence and the processing technology assessment result sequence, combined with the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence, synchronous abnormality judgment is performed on the equipment operation node sub-chain and the process node sub-chain to obtain the full-process associated abnormality assessment space.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] In response to the deficiencies of the prior art, the present invention realizes full-dimensional real-time synchronous monitoring of equipment operation status and processing technology information by constructing a dual monitoring node chain and a hierarchical abnormality association evaluation node chain. Combined with the multi-level threshold discrimination of the operation hierarchical evaluation indicator set and the hierarchical process evaluation indicator set, it effectively distinguishes complex scenarios such as single equipment abnormalities, process independent abnormalities and cross-node collaborative abnormalities; based on the Bayesian causal reasoning model and the collaborative causal process connection, it accurately locates the abnormal propagation path and correlation contribution of equipment operation nodes and process nodes, and enhances the depth and accuracy of abnormality tracing; dynamically matches the historical abnormality strategy library through a distributed algorithm, generates differentiated processing strategies for different abnormality types, combines the two-way operation abnormality reasoning connection with the reverse deduction of process parameter deviation, realizes closed-loop dynamic compensation and correction of equipment control parameters and process parameters, improves the real-time and adaptability of abnormality processing, and reduces the risk of global process chain fluctuations caused by local abnormalities. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of a seamless pipe fitting according to embodiment 1 of the present invention;

[0048] Figure 2 This is a flow chart of a method for detecting and handling anomalies in a smart factory production process according to embodiment 1 of the present invention;

[0049] Figure 3 This is a logical architecture diagram of a method for detecting and handling anomalies in a smart factory production process according to embodiment 1 of the present invention;

[0050] Figure 4 This is a diagram of the production process flow chain abnormality monitoring architecture of Example 1 of the present invention. DETAILED DESCRIPTION

[0051] Example 1

[0052] As high-end manufacturing industries increase their quality requirements for high-precision steel pipes and pipe fittings, lean manufacturing-based production processes urgently need to achieve refined control over the entire life cycle. In device production control, for example Figure 1 The seamless pipe fitting process involves a complex process involving more than ten steps, including blanking, pickling, and solution treatment. Slight deviations in process parameters at each stage (such as fluctuations in acid concentration or dimensional tolerance violations) can trigger chain reactions across multiple steps, leading to batch scrapping or production line paralysis. However, existing anomaly detection technologies suffer from isolated detection, delayed response, and difficulty locating the root causes of anomalies. Tracing the root causes of anomalies relies on manual judgment and static threshold limitations, leading to frequent false alarms or missed detections, impacting production continuity. To address this issue, please refer to Figure 2 、 Figure 3 The present invention provides an embodiment of a method for detecting and processing abnormalities in a production process of a smart factory, comprising the following steps:

[0053] S1. Based on the preset production process chain, a dual monitoring node chain and a hierarchical abnormality correlation evaluation node chain are constructed;

[0054] Furthermore, the hierarchical anomaly association evaluation node chain in this embodiment includes an operation status evaluation subchain and a full-process process anomaly evaluation subchain;

[0055] The production process chain includes the process node sub-chain and the equipment operation node sub-chain; the dual monitoring node chain includes the equipment operation monitoring sub-chain and the process monitoring sub-chain;

[0056] Furthermore, each processing node in the process node sub-chain in this embodiment stores the processing parameters of the corresponding node;

[0057] Furthermore, the specific processing parameters in this embodiment are specifically set by those skilled in the art according to the specific processing procedures; the device control parameters of the device operation node sub-chain are also similar and are specifically set by those skilled in the art according to the machine model and specific equipment;

[0058] Each operation node in the equipment operation node subchain stores the equipment control parameters for the corresponding processing equipment to execute the processing technology;

[0059] The equipment operation monitoring subchain corresponds one-to-one with the nodes in the equipment operation node subchain; the process node subchain corresponds one-to-one with the nodes in the process monitoring subchain;

[0060] Each monitoring node in the equipment operation monitoring subchain corresponds one-to-one to an operation evaluation node in the operation status evaluation subchain.

[0061] Furthermore, each process monitoring node in the process monitoring subchain in this embodiment corresponds one-to-one to each process assessment node in the full-process process anomaly assessment subchain;

[0062] Each operation evaluation node in the operation status evaluation subchain is connected to the process evaluation node in the full process process abnormality evaluation subchain through a one-to-many backward mapping;

[0063] Each operation evaluation node is configured with a set of hierarchical operation evaluation indicators for evaluating the operation status of the corresponding processing equipment; each process evaluation node is configured with a set of hierarchical process evaluation indicators for evaluating the corresponding processing technology;

[0064] The operation hierarchical evaluation index set and the hierarchical process evaluation index set are respectively stored in a preset hierarchical operation index library and a hierarchical process index library, and a one-to-many directed connection is performed between the hierarchical process index library and the hierarchical operation index library through a preset directed association connection.

[0065] Furthermore, in this embodiment, the specific process of performing a one-to-many directed connection between the hierarchical process index library and the hierarchical operation index library through a preset directed association connection includes:

[0066] According to all the hierarchical process evaluation indicator sets and all the operation hierarchical evaluation indicator sets, the correlation analysis method is used to obtain the correlation influence degree between the operation hierarchical evaluation indicator corresponding to each equipment operation node and the corresponding hierarchical process evaluation indicator in the subsequent processing process node. Based on the correlation influence degree between the operation hierarchical evaluation indicator corresponding to the equipment operation node and the corresponding hierarchical process evaluation indicator in the subsequent processing process node, a one-to-many directed connection is constructed between each operation hierarchical evaluation indicator and multiple hierarchical process evaluation indicators.

[0067] Furthermore, in the seamless steel pipe production process, in order to achieve precise control of the production process and quality optimization, the hierarchical operation evaluation index set and the hierarchical process evaluation index set are stored in the preset hierarchical operation index library and hierarchical process index library respectively, and a one-to-many directed connection is performed through a preset directed association connection;

[0068] Furthermore, to better illustrate the process of constructing a process node subchain, this embodiment uses the seamless pipe production process as an example. The process node subchain includes processing nodes corresponding to more than ten processes, including blanking, pickling, and solutionizing. Each processing node stores the processing parameters for the corresponding process. For example, in the pickling process, parameters such as acid concentration, pickling time, and temperature are recorded; in the blanking process, parameters such as pipe size and cutting accuracy are key process parameters. These parameters are the basis for ensuring the normal operation of each process and guaranteeing product quality, and they determine the specific operating standards for each process.

[0069] During the construction of the equipment operation node sub-chain: each operation node in the equipment operation node sub-chain corresponds to the equipment that performs each processing technology. Each operation node saves the equipment control parameters for the corresponding processing equipment to perform the processing technology. For example, in the pickling equipment operation node, the equipment's stirring speed, acid circulation flow rate and other control parameters are recorded and stored; for solution treatment equipment, parameters such as heating temperature control, heating time control and cooling rate control are information that the equipment operation node needs to save; these parameters directly affect the operating status of the equipment, and thus affect the execution effect of the process.

[0070] Furthermore, the process of constructing a dual monitoring node chain in this embodiment includes:

[0071] The equipment operation monitoring subchain in this embodiment corresponds one-to-one to the nodes in the equipment operation node subchain; taking seamless pipe production as an example, various sensors, such as temperature sensors, pressure sensors, flow sensors, etc., are installed on each device; for pickling equipment, the acid temperature is monitored by a temperature sensor, the pressure sensor monitors the pressure of the acid circulation system, and the flow sensor monitors the flow of the acid; these sensors collect various data during the operation of the equipment in real time, forming the data source of the equipment operation monitoring subchain; through real-time monitoring of these data, abnormal conditions during the operation of the equipment can be discovered in a timely manner, such as excessive temperature may mean that there is a problem with the heat dissipation of the equipment, and abnormal pressure fluctuations may indicate pipe blockage or abnormal operation of the pump.

[0072] Furthermore, the process monitoring subchain in this embodiment corresponds one-to-one to the nodes in the process node subchain; monitoring points are set in each process link to monitor the key indicators of different processes; in the pickling process, in addition to monitoring the temperature, concentration and other equipment operation-related parameters of the acid solution, it is also necessary to monitor process indicators such as the degree of corrosion and cleanliness of the pipe surface; in the blanking process, the size of the pipe after cutting is measured in real time to monitor whether it meets the specified dimensional tolerance range. These process monitoring data can directly reflect the effect of process execution and help determine whether the process is carried out within the normal range. It should be further explained that in this embodiment, monitoring points are set in each process link, and the monitoring equipment used at the monitoring points is specifically set by those skilled in the art according to the monitoring process requirements; for example, in the blanking process, the size of the pipe after cutting is monitored by a laser caliper to scan the outer diameter of the pipe, combined with a contact grating ruler to measure the length, and the data processing system is used to calculate in real time whether the dimensional deviation is within the tolerance range; for the form and position tolerances such as the verticality of the pipe, a three-coordinate measuring machine is used for sampling detection, or a visual inspection system is used to capture an image of the pipe end face and analyze the edge verticality deviation.

[0073] Furthermore, in this embodiment, each monitoring node in the equipment operation monitoring subchain corresponds one-to-one with an operation evaluation node in the operation status evaluation subchain. The operation evaluation node receives monitoring data from the equipment operation monitoring subchain and uses a comprehensive fuzzy algorithm combined with a Bayesian function to evaluate the equipment's operation status. For example, a parameter model of the equipment's normal operation is established using historical data and a comprehensive fuzzy algorithm combined with a Bayesian function. When the monitoring data deviates from the normal model range, the evaluation node can determine that the equipment is in an abnormal state. For pickling equipment, if the acid solution temperature remains above the normal range and the fluctuation amplitude exceeds a set threshold, the operation evaluation node will determine that the equipment's operation status is abnormal. Simultaneously, by combining information such as the equipment's operating time and maintenance records, the operation status of the equipment can be more accurately assessed and potential failures predicted.

[0074] Furthermore, each process monitoring node in the process monitoring subchain in this embodiment corresponds one-to-one to each process evaluation node in the full-process process anomaly evaluation subchain; the full-process process anomaly evaluation subchain integrates the process data transmitted by each process monitoring node, combines the requirements and standards of the entire production process, and determines whether there is an anomaly in the process; for example, in the production of seamless pipe fittings, it is necessary to consider not only the process parameters of a single process, but also the relationship between the processes. If the surface cleanliness of the pipe after pickling does not meet the standard, it may affect the effect of the subsequent solution treatment, and thus affect the product quality. By comprehensively analyzing the data of multiple process monitoring points, it is possible to more comprehensively determine whether the process is normal.

[0075] The connection between the operation status assessment subchain and the full-process process anomaly assessment subchain: Each operation assessment node in the operation status assessment subchain is connected to the process assessment node in the full-process process anomaly assessment subchain through a one-to-many backward mapping. This means that when the full-process process anomaly assessment subchain finds an anomaly in a certain process link, the operation status information of the equipment related to that process link can be quickly located through this mapping relationship. For example, if a problem is found with the quality of pipe fittings after solution treatment, the full-process process anomaly assessment subchain can view the operation status assessment information of the solution treatment equipment and the equipment in the previous related processes through the mapping relationship.

[0076] S2. Real-time monitoring of the production process chain based on the dual monitoring node chain to obtain equipment operating status information and production process information of the equipment to be processed;

[0077] S3. Based on the equipment operation status information and the production process information of the equipment to be processed, combined with the hierarchical abnormality association evaluation node chain, obtain the full-process associated abnormality evaluation space;

[0078] S4. Based on the full-process associated anomaly evaluation space, an anomaly handling strategy library is constructed through a distributed algorithm combined with historical anomaly strategy index information to obtain a differential anomaly handling strategy, and the differential anomaly handling strategy is fed back to the dual monitoring node chain in real time to provide real-time warning and processing for the anomalies monitored and located in the production process chain.

[0079] Furthermore, the exception handling strategy library in this embodiment is connected by association mapping through the corresponding association between each exception strategy keyword in the exception handling strategy library and the operation indicator keyword in the hierarchical operation indicator library.

[0080] Furthermore, the exception handling strategy library in this embodiment quickly indexes the exception handling strategies associated with the hierarchical operation indicator library through an association analysis algorithm combined with an association mapping connection to obtain differential exception handling strategies.

[0081] Furthermore, the construction process of the production process chain in this embodiment includes:

[0082] Obtain the full-process production process information of each device to be processed and combine it with the pre-trained production process analysis algorithm to obtain the continuous processing process node sequence and processing order information corresponding to each device to be processed, as well as the required processing equipment sequence and processing equipment operation interaction information, as well as the mapping information and associated causal information of each processing technology and processing equipment;

[0083] Furthermore, the production process parsing algorithm in this embodiment is constructed by using a pre-trained Chinese BERT model.

[0084] Based on the continuous processing node sequence and processing order information corresponding to each device to be processed, a loosely coupled algorithm is used to construct a directed loosely coupled connection between each processing node corresponding to each device to be processed;

[0085] Based on the continuous processing process node sequence and the directed loosely coupled connection, obtaining a process node sub-chain corresponding to each device to be processed;

[0086] At the same time, the equipment operation node sequence is constructed based on the required processing equipment sequence, and the interactive information of each processing equipment operation is used to build a two-way operation abnormality reasoning connection through the graph neural network;

[0087] Based on the device operation node sequence combined with the bidirectional operation abnormality reasoning connection, a device operation node sub-chain is obtained.

[0088] Assume that the current process node sub-chain contains M continuous processing nodes, the equipment operation node sub-chain contains N equipment operation nodes, and the process parameters corresponding to each processing node are only processed and run on one equipment operation node;

[0089] Based on the assumption, the processing node corresponding to the process parameters running on each equipment operation node is set as the main processing node of the equipment operation node, and the remaining continuous processing nodes after the main processing node constitute the causal relationship process node sequence of the current equipment operation node;

[0090] Based on the main causal correlation information between the operating status and processing results of each equipment operation node and the corresponding main processing node, the main correlation degree between each equipment operation node and the corresponding main processing node is constructed through the autocorrelation algorithm, and the main connection relationship between the corresponding equipment operation node and the main processing node is constructed using the obtained main correlation degree;

[0091] At the same time, based on the collaborative operation causal information between each equipment operation node and the corresponding causal correlation process node sequence, the collaborative causal correlation degree of each equipment operation node for each processing node in the corresponding causal correlation process node sequence is obtained through the cross-correlation algorithm;

[0092] Based on the collaborative causal correlation degree of each equipment operation node to each processing node of the corresponding causal correlation process node sequence, a collaborative causal process connection is constructed between the equipment operation node and the corresponding causal correlation process node sequence;

[0093] Connect the nodes corresponding to the process node subchain and the equipment operation node subchain based on the main connection relationship and the collaborative causal process connection to obtain a production process flow chain;

[0094] For a more detailed explanation of the construction process of the production process chain, please refer to Figure 4 , where A represents the process node sub-chain and B represents the equipment operation node sub-chain. The process node sub-chain contains M continuous processing nodes, and the equipment operation node sub-chain contains N equipment operation nodes. The corresponding unidirectional arrow in A is a directed loosely coupled connection, which is used to use blockchain technology for secure transmission and interaction of process parameters; the bidirectional arrow in B is a bidirectional operation anomaly reasoning connection, which is used to trace the anomaly or fault forward or backward and locate the anomaly or fault.

[0095] Furthermore, in this embodiment, 1 to M-1 corresponding to between A and B is the main connection relationship and collaborative causal process connection between the equipment operation node 1 and the processing technology nodes 1 to M, wherein in the corresponding connection between A and B, 1 represents the main connection, and 2 to M-1 is the collaborative causal process connection between the equipment operation node 1 and the corresponding causal association process node sequence, which is used to forward trace the production process deviation of the processing equipment when the processing is abnormal and the equipment does not detect the abnormality, and accurately correct the deviation on the cumulative path of the process deviation after tracing.

[0096] A causal reasoning model is constructed based on the Bayesian algorithm and configured into the bidirectional operation abnormality reasoning connection and the collaborative causal process connection to obtain a production process flow chain with causal reasoning.

[0097] Furthermore, the process of constructing the dual monitoring node chain in this embodiment includes:

[0098] Based on the architecture mapping of the device operation node subchain, a device operation monitoring subchain with the same architecture is obtained, and monitoring equipment is configured according to the corresponding operation attributes of each device operation node to obtain the device operation parameter space corresponding to the device operation node subchain;

[0099] The device operation parameter space includes a single operation parameter sequence corresponding to each device operation node and an interactive operation parameter sequence corresponding between each device operation node;

[0100] At the same time, based on the architecture mapping of the process node sub-chain, a process monitoring sub-chain with the same architecture is obtained, and the corresponding monitoring equipment is configured according to the processing technology information to obtain the processing technology monitoring information space corresponding to the equipment operation node sub-chain.

[0101] Furthermore, the process of constructing the hierarchical abnormality association evaluation node chain in this embodiment includes:

[0102] Based on the architecture mapping of the equipment operation node subchain, the operation status evaluation subchain of the same architecture is obtained, and based on the architecture mapping of the process node subchain, the full-process process anomaly evaluation subchain of the same architecture is obtained;

[0103] The main connection relationship and the collaborative causal process connection in the production process flow chain with causal reasoning are mapped to the corresponding nodes in the operation status evaluation subchain and the full-process process anomaly evaluation subchain, and a one-to-many backward mapping connection between the operation status evaluation subchain and the full-process process anomaly evaluation subchain is constructed to obtain a hierarchical anomaly association evaluation node chain.

[0104] Furthermore, the process of obtaining the differential exception handling strategy in this embodiment includes:

[0105] Based on the dual monitoring node chain, the equipment operation parameter space and the processing technology monitoring information space are monitored in real time, and synchronously mapped to the corresponding operation status assessment subchain and full-process process anomaly assessment subchain in the hierarchical anomaly association assessment node chain;

[0106] Based on the hierarchical operation evaluation indicator set configured for each operation evaluation node in the operation status evaluation subchain and the full-process process abnormality evaluation subchain, and the hierarchical process evaluation indicator set configured for each process evaluation node, the operation status and processing process results of each operation evaluation node and the corresponding main processing process node are synchronously evaluated to obtain an operation status evaluation result sequence and a processing process evaluation result sequence;

[0107] Set the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence, and based on the operating status assessment result sequence and the processing technology assessment result sequence combined with the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence, perform synchronous abnormality judgment on the equipment operation node sub-chain and the process node sub-chain to obtain the full-process associated abnormality assessment space.

[0108] Based on the full-process associated anomaly evaluation space, if at least one pair of equipment operation nodes and corresponding main processing nodes in the equipment operation node subchain and the process node subchain is in an abnormal state, then according to the anomaly evaluation scores of the located abnormal equipment operation nodes and corresponding main processing nodes, as well as the processing parameters and equipment control parameters, combined with the anomaly handling strategy library, a first differential anomaly handling strategy is obtained;

[0109] It should be further explained that one implementation of obtaining the first difference exception handling strategy in this embodiment is:

[0110] Based on the full-process associated anomaly assessment space, locate the equipment operation nodes and corresponding main processing nodes in abnormal states, and extract the anomaly assessment scores, processing parameter deviation values, and equipment control parameter deviation values ​​of the abnormal nodes;

[0111] Extracting operation indicator keywords from the abnormality assessment score, the processing parameter deviation value, and the equipment control parameter deviation value using an entity-relationship extraction algorithm, wherein the operation indicator keywords include equipment type, process type, parameter name, and deviation range;

[0112] The extracted operation indicator keywords are matched with the abnormal strategy keywords preset in the abnormal handling strategy library. The candidate abnormal handling strategies with a correlation degree ≥ the threshold are screened out through the association analysis algorithm.

[0113] It should be further explained that, in this embodiment, the exception handling policy library pre-stores a number of exception handling policy records, each record including a policy ID and an exception policy keyword;

[0114] The abnormal nodes are graded according to their abnormality assessment scores combined with the preset abnormality level intervals, and the candidate abnormality handling strategies are prioritized according to the grading results. The parameters of the ranked candidate strategies are adapted and adjusted based on the equipment operation and maintenance history records and process execution history data to generate the first differential abnormality handling strategy.

[0115] It should be further explained that this embodiment adjusts the parameters in the candidate strategy based on the operation and maintenance records of similar equipment within a preset time period and historical process execution data, such as the optimal correction value under the same deviation;

[0116] It should be further explained that the first differential exception handling strategy generated in this embodiment should include but not be limited to: target node ID, correction parameter name, parameter values ​​before and after correction, execution time limit, etc.;

[0117] The first difference exception handling strategy is fed back to the dual monitoring node chain in real time, triggering the corresponding exception warning mechanism and driving the device to perform parameter correction operations.

[0118] If only the processing node is in an abnormal state at the current moment, and the equipment operation node is evaluated as normal, then based on the equipment operation node corresponding to the abnormal processing node, the processing abnormality is located by performing dual forward reasoning through the bidirectional operation abnormality reasoning connection and the collaborative causal process connection built into the production process flow chain with causal reasoning, and a deviation accumulation equipment operation node sequence is obtained;

[0119] Based on the deviation accumulation equipment operation node, the processing technology evaluation result corresponding to the main processing technology node is obtained;

[0120] Based on the processing technology evaluation results corresponding to the main processing technology nodes, the process parameter deviation value corresponding to each main processing technology node that meets the process requirements is obtained;

[0121] Based on the process parameter deviation value that meets the process requirements corresponding to each main processing process node, the equipment control parameter deviation value of the equipment operation node corresponding to each main processing process node is obtained;

[0122] Based on the equipment control parameter deviation value corresponding to the deviation accumulation equipment operation node and combined with the exception handling strategy library, a second difference exception handling strategy is obtained, and based on the second difference exception handling strategy, the process parameter deviation value corresponding to the deviation accumulation equipment operation node is corrected, so that the processing technology evaluation result corresponding to each processing technology node meets the corresponding processing technology exception evaluation threshold in real time.

[0123] It should be further explained that the second difference exception handling strategy in this embodiment is specifically implemented as follows:

[0124] Based on the unique identification ID of the abnormal process node, its corresponding equipment operation node is determined. Through the bidirectional operation abnormality reasoning connection and the collaborative causal process connection of the built-in Bayesian causal reasoning model in the production process flow chain, forward reasoning is performed with the abnormal process node as the starting point, and the deviation transfer coefficient of the equipment operation node to the subsequent process nodes is calculated. The equipment operation nodes with a deviation transfer coefficient ≥ the preset threshold are screened out, and the deviation-accumulated equipment operation node sequence is formed by arranging them in descending order according to the deviation impact weight, where the deviation transfer coefficient is obtained through the cross-correlation analysis of historical process deviation data and equipment control parameters.

[0125] Based on the main processing node identification of each node in the deviation accumulation equipment operation node sequence, the corresponding process evaluation result in the full-process process anomaly evaluation sub-chain is called, and the difference between the current process parameter measured value and the process standard value of the main processing process node is extracted. Combined with the allowable deviation threshold of the main processing process node in the hierarchical process indicator library, the process parameter deviation value that meets the process requirements is calculated through the deviation compensation algorithm. The deviation compensation algorithm needs to take into account the parameter sensitivity coefficient of subsequent causally related process nodes to ensure that the calculation result is within the process tolerance range.

[0126] According to the parameter mapping table preset in the main connection relationship between the process node sub-chain and the equipment operation node sub-chain, the determined process parameter deviation value is converted into the equipment control parameter deviation value of the corresponding equipment operation node. The conversion process requires the introduction of the equipment characteristic curve correction coefficient, which is obtained by comparing and analyzing the equipment factory parameters and the actual operation calibration data. For example, when converting the acid concentration deviation value of the pickling process into the acid circulation flow deviation value of the pickling equipment, dynamic correction is required according to the acid concentration-flow response curve.

[0127] The obtained equipment control parameter deviation value is used as a keyword, and cosine similarity matching is performed with the abnormal strategy keywords stored in the abnormal handling strategy library to screen out the candidate strategy with the highest similarity. Combined with the model parameters and historical correction records of the deviation accumulation equipment operation node, the control parameter adjustment range in the candidate strategy is adapted and optimized to generate a second differential abnormal handling strategy containing the correction object ID, control parameter name, values ​​before and after adjustment, execution order and verification threshold. The strategy is sent to the corresponding equipment operation node through the dual monitoring node chain to drive the equipment to perform parameter correction operations, and the corrected process parameters are collected in real time through the process monitoring sub-chain until the process evaluation result meets the corresponding processing technology abnormality evaluation threshold.

[0128] In summary, this embodiment realizes a multi-dimensional abnormality monitoring and precise control system in the seamless steel pipe production scenario by constructing a dual-chain collaborative architecture of the process node sub-chain and the equipment operation node sub-chain, combining the dynamic perception of the dual monitoring node chain and the causal reasoning mechanism of the hierarchical abnormality association evaluation node chain; specifically: first, the equipment operation monitoring sub-chain captures the high-frequency time series characteristics of the equipment control parameters in real time through the multi-dimensional sensor network, and establishes the equipment health baseline model with the help of the Bayesian-fuzzy comprehensive algorithm in the operation status evaluation sub-chain. This step enables the equipment's hidden abnormalities to be detected by fusing the time-frequency domain characteristics of the equipment operation parameters with the knowledge embedding of historical maintenance data. Breaking through the limitations of traditional threshold detection, the feature matching model can still identify the equipment operation evaluation score when it does not trigger an alarm. This evaluation mechanism based on deep characterization of multi-source data significantly improves the sensitivity of equipment status monitoring and the ability to suppress false alarms. Secondly, the process monitoring sub-chain constructs a full-process process anomaly evaluation network through the parameter constraint relationship between the cross-process quality inspection node and the process node sub-chain. When the surface roughness Ra value is detected to be excessive, the preset correlation influence degree in the hierarchical process indicator library is used to quickly locate the pickling process equipment node through backward mapping connection, and combined with the forward propagation simulation and reverse tracing analysis of the bidirectional graph neural network, the process anomaly and equipment are realized. The causal decoupling of latent faults, this cross-layer associative reasoning mechanism effectively solves the problem of monitoring blind spots when the equipment is normal but the process is abnormal; furthermore, the cascade design of loosely coupled connections and collaborative causal process connections in the production process chain enables the process chain deviation caused by the pickling concentration deviation to be gradually eliminated among multiple processes according to the influence weight through a distributed correction strategy, avoiding the risk of secondary abnormalities caused by traditional single-point parameter over-adjustment. The realization of this technical effect depends on the accurate modeling of the process parameter transfer function and the equipment control response characteristics; in addition, the blockchain-enhanced multi-level signature verification mechanism and the smart contract-driven parameter collaborative adjustment transform the process chain adjustment process into a legally binding mechanism. The digital certificate of effectiveness is automatically generated through the feature precipitation and rule generation of the case library, and a knowledge closed-loop evolution system is constructed, which enables the system to continuously adapt to new abnormal patterns. This technological integration not only ensures the auditability and compliance of process adjustments, but also realizes the digital inheritance of empirical knowledge through machine learning mechanisms; finally, the dynamic network topology and evidence weight distribution strategy of the Bayesian causal reasoning engine, when detecting cold-rolled size deviations, guides the correction strategy to prioritize triggering laser scanning re-inspection and synchronously adjust the pickling spray angle through reverse probability calculation and Monte Carlo simulation. This decision-making mechanism based on probability reasoning significantly improves the processing efficiency of complex abnormal scenarios.The synergistic effect of the above-mentioned technical means essentially builds a closed-loop intelligent manufacturing system for data perception, feature extraction, causal reasoning, dynamic correction, and knowledge evolution. The implicit technological breakthrough lies in: through the digital twin modeling of the process chain and the autonomous learning of the causal association network, the traditional passive quality control based on the rule engine is upgraded to a cognitive intelligent system with predictive control capabilities, ensuring the production quality of high-precision pipe fittings.

[0129] Example 2

[0130] The present application discloses an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements an abnormality detection and processing method for a smart factory production process when executing the computer program.

[0131] The present application also discloses a computer-readable storage medium having computer instructions stored thereon, which, when the computer instructions are executed, executes a method for detecting and processing anomalies in a smart factory production process.

[0132] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also change, modify, replace and modify the above-mentioned embodiments without departing from the scope of protection of the purpose of the present invention and the claims, and all of these are protected by the present invention.

Claims

1. A method for detecting and processing abnormalities in the production process of a smart factory, characterized in that: include: Based on the preset production process chain, a dual monitoring node chain and a hierarchical abnormality correlation evaluation node chain are constructed; The hierarchical abnormality association evaluation node chain includes an operation status evaluation subchain and a full-process process abnormality evaluation subchain; Based on the dual monitoring node chain, the production process chain is monitored in real time to obtain equipment operation status information and production process information of the equipment to be processed; Based on the equipment operation status information and the production process information of the equipment to be processed, combined with the hierarchical abnormality association evaluation node chain, the whole process associated abnormality evaluation space is obtained; Based on the full-process associated anomaly evaluation space, an anomaly handling strategy library is constructed through a distributed algorithm combined with historical anomaly strategy index information to obtain a differential anomaly handling strategy, and the differential anomaly handling strategy is fed back to the dual monitoring node chain in real time to provide real-time warning and processing for the anomalies monitored and located in the production process chain.

2. The method for detecting and handling abnormalities in a production process of a smart factory according to claim 1, wherein: The production process chain includes a process node sub-chain and an equipment operation node sub-chain; the dual monitoring node chain includes an equipment operation monitoring sub-chain and a process monitoring sub-chain; Each processing node in the process node subchain stores the processing parameters of the corresponding node; each operation node in the equipment operation node subchain stores the equipment control parameters for the corresponding processing equipment to execute the processing technology; The equipment operation monitoring subchain has a one-to-one correspondence with the nodes in the equipment operation node subchain; the process node subchain has a one-to-one correspondence with the nodes in the process monitoring subchain; Each monitoring node in the equipment operation monitoring subchain corresponds one-to-one to an operation evaluation node in the operation status evaluation subchain.

3. The method for detecting and handling abnormalities in a production process of a smart factory according to claim 2, wherein: Each process monitoring node in the process monitoring subchain corresponds one-to-one to each process assessment node in the full-process process abnormality assessment subchain; Each operation evaluation node in the operation status evaluation subchain is connected to the process evaluation node in the full process process abnormality evaluation subchain in a one-to-many backward mapping manner; Each operation evaluation node is configured with a hierarchical operation evaluation index set for evaluating the operation status of the corresponding processing equipment; each process evaluation node is configured with a hierarchical process evaluation index set for evaluating the corresponding processing technology; The operation hierarchical evaluation index set and the hierarchical process evaluation index set are respectively stored in a preset hierarchical operation index library and a hierarchical process index library, and a one-to-many directed connection is performed on the hierarchical process index library and the hierarchical operation index library through a preset directed association connection.

4. The method for detecting and handling abnormalities in a production process of a smart factory according to claim 3, wherein: The construction process of the production process chain includes: Obtain the full-process production process information of each device to be processed and combine it with the pre-trained production process analysis algorithm to obtain the continuous processing process node sequence and processing order information corresponding to each device to be processed, as well as the required processing equipment sequence and processing equipment operation interaction information, as well as the mapping information and associated causal information of each processing technology and processing equipment; Based on the continuous processing node sequence and processing order information corresponding to each device to be processed, a loosely coupled algorithm is used to construct a directed loosely coupled connection between each processing node corresponding to each device to be processed; Based on the continuous processing process node sequence and the directed loosely coupled connection, obtaining a process node sub-chain corresponding to each device to be processed; At the same time, the equipment operation node sequence is constructed based on the required processing equipment sequence, and the interactive information of each processing equipment operation is used to build a two-way operation abnormality reasoning connection through the graph neural network; Based on the device operation node sequence combined with the bidirectional operation abnormality reasoning connection, a device operation node sub-chain is obtained.

5. The method for detecting and handling abnormalities in a production process of a smart factory according to claim 4, wherein: The construction process of the production process chain also includes: Assume that the current process node sub-chain contains M continuous processing nodes, the equipment operation node sub-chain contains N equipment operation nodes, and the process parameters corresponding to each processing node are only processed and run on one equipment operation node; Based on the assumption, the processing node corresponding to the process parameters running on each equipment operation node is set as the main processing node of the equipment operation node, and the remaining continuous processing nodes after the main processing node constitute the causal relationship process node sequence of the current equipment operation node; Based on the main causal correlation information between the operating status and processing results of each equipment operation node and the corresponding main processing node, the main correlation degree between each equipment operation node and the corresponding main processing node is constructed through the autocorrelation algorithm, and the main connection relationship between the corresponding equipment operation node and the main processing node is constructed using the obtained main correlation degree; At the same time, based on the collaborative operation causal information between each equipment operation node and the corresponding causal process node sequence, the collaborative causal correlation degree of each equipment operation node for each processing node in the corresponding causal process node sequence is obtained through the cross-correlation algorithm.

6. The method for detecting and handling abnormalities in a smart factory production process according to claim 5, wherein: The construction process of the production process chain also includes: Based on the collaborative causal correlation degree of each equipment operation node to each processing node of the corresponding causal correlation process node sequence, a collaborative causal process connection is constructed between the equipment operation node and the corresponding causal correlation process node sequence; Connect the nodes corresponding to the process node subchain and the equipment operation node subchain based on the main connection relationship and the collaborative causal process connection to obtain a production process flow chain; A causal reasoning model is constructed based on the Bayesian algorithm and configured into the bidirectional operation abnormality reasoning connection and the collaborative causal process connection to obtain a production process flow chain with causal reasoning.

7. The method for detecting and handling abnormalities in a smart factory production process according to claim 6, wherein: The construction process of the dual monitoring node chain includes: Based on the architecture mapping of the device operation node subchain, a device operation monitoring subchain with the same architecture is obtained, and monitoring equipment is configured according to the corresponding operation attributes of each device operation node to obtain the device operation parameter space corresponding to the device operation node subchain; The device operation parameter space includes a single operation parameter sequence corresponding to each device operation node and an interactive operation parameter sequence corresponding between each device operation node; At the same time, based on the architecture mapping of the process node sub-chain, a process monitoring sub-chain with the same architecture is obtained, and the corresponding monitoring equipment is configured according to the processing technology information to obtain the processing technology monitoring information space corresponding to the equipment operation node sub-chain.

8. The method for detecting and handling abnormalities in a smart factory production process according to claim 7, wherein: The process of constructing the hierarchical abnormal correlation evaluation node chain includes: Based on the architecture mapping of the equipment operation node subchain, the operation status evaluation subchain of the same architecture is obtained, and based on the architecture mapping of the process node subchain, the full-process process anomaly evaluation subchain of the same architecture is obtained; The main connection relationship and the collaborative causal process connection in the production process flow chain with causal reasoning are mapped to the corresponding nodes in the operation status evaluation subchain and the full-process process anomaly evaluation subchain, and a one-to-many backward mapping connection between the operation status evaluation subchain and the full-process process anomaly evaluation subchain is constructed to obtain a hierarchical anomaly association evaluation node chain.

9. The method for detecting and handling abnormalities in a smart factory production process according to claim 8, wherein: The process of obtaining the differential exception handling strategy includes: Based on the dual monitoring node chain, the equipment operation parameter space and the processing technology monitoring information space are monitored in real time, and synchronously mapped to the corresponding operation status assessment subchain and full-process process anomaly assessment subchain in the hierarchical anomaly association assessment node chain; Based on the hierarchical operation evaluation indicator set configured for each operation evaluation node in the operation status evaluation subchain and the full-process process abnormality evaluation subchain, and the hierarchical process evaluation indicator set configured for each process evaluation node, the operation status and processing process results of each operation evaluation node and the corresponding main processing process node are synchronously evaluated to obtain an operation status evaluation result sequence and a processing process evaluation result sequence; Set the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence. Based on the operating status assessment result sequence and the processing technology assessment result sequence, combined with the operating status abnormality assessment threshold sequence and the processing technology abnormality assessment threshold sequence, synchronous abnormality judgment is performed on the equipment operation node sub-chain and the process node sub-chain to obtain the full-process associated abnormality assessment space.

Images