A production equipment visual operation and maintenance method and system based on digital twinning

Abstract

The application relates to the technical field of industrial operation and maintenance, and discloses a production equipment visual operation and maintenance method and system based on digital twinning, which comprises the following steps: structurally recombining real-time operation data to obtain operation data base; virtually and virtually synchronously binding a three-dimensional geometric skeleton and the operation data base to obtain an initial digital twinning body; anchoring the operation parameter nodes in the initial digital twinning body to historical fault records of a target production equipment to obtain an associated digital twinning body; performing real-time abnormal threshold detection and intelligent traceability deduction on the fault parameter nodes in the associated digital twinning body to obtain a preliminary diagnosis report; performing hierarchical coupling rendering on the preliminary diagnosis report and the associated digital twinning body to obtain a visual operation and maintenance interface; and performing visual operation chain compilation on the interactive operation behavior of the visual operation and maintenance interface to obtain precise regulation and control operation instructions; and the application can improve the efficiency of the production equipment visual operation and maintenance based on digital twinning.

Description

Technical Field

[0001] This invention relates to the field of industrial operation and maintenance technology, and in particular to a method and system for visual operation and maintenance of production equipment based on digital twins. Background Technology

[0002] In existing production equipment visualization and maintenance technologies, there is a lack of effective means for processing real-time equipment operation data through multi-source heterogeneous fusion and tensor structured weaving. The data has not achieved isomorphic mapping with the equipment's logical topology, and the construction of the operation data base lacks support from spatiotemporal characteristics and topological correlations, resulting in insufficient data effectiveness and correlation. At the same time, in the digital twin construction stage, the construction of the three-dimensional geometric skeleton does not rely on the topological constraints of the equipment's physical structure for parametric surface reconstruction, and the virtual-real synchronous binding with the operation data base also lacks accurate mapping relationship matrix support. The accuracy of virtual-real fusion is low, and it is impossible to form an initial digital twin that can truly reflect the equipment's operating status.

[0003] Current technologies in equipment fault operation and maintenance lack scientific fault feature anchoring between digital twin operating parameter nodes and historical fault records, and lack confidence matrix enhancement and filtering mechanisms, resulting in insufficient accuracy of fault association marking. Anomaly detection of fault parameter nodes only involves basic threshold judgments, lacking effective causal link tracing and fault propagation topology inversion capabilities. This leads to low efficiency and accuracy in root cause tracing. Furthermore, the fusion and rendering of diagnostic reports and digital twins lacks hierarchical and targeted matching design, resulting in low information transmission efficiency in the visualized operation and maintenance interface. Simultaneously, the parsing of interface interactions remains superficial, failing to efficiently translate operational behaviors into precise equipment control commands. The overall intelligence and visualization level of the operation and maintenance process is insufficient to meet the needs of production equipment operation and maintenance. Therefore, improving the efficiency and accuracy of visualized operation and maintenance of production equipment based on digital twins has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for visual operation and maintenance of production equipment based on digital twins, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a visual operation and maintenance method for production equipment based on digital twins, comprising: P1. Collect real-time operating data of the target production equipment and perform structured reorganization on the real-time operating data to obtain the operating data base of the target production equipment; P2. Based on the physical structure of the target production equipment, construct a three-dimensional geometric skeleton of the target production equipment, and bind the three-dimensional geometric skeleton to the operating data base in a virtual-real synchronous manner to obtain the initial digital twin of the target production equipment; P3. Anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment; P4. Real-time abnormal threshold detection is performed on the fault parameter nodes in the associated digital twin, and intelligent source tracing and deduction are performed on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment. P5. Perform hierarchical coupling rendering of the preliminary diagnostic report and the associated digital twin to obtain the visual operation and maintenance interface of the target production equipment; P6. Compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain the precise control and operation and maintenance instructions for the target production equipment.

[0006] In a preferred embodiment, the step of collecting real-time operating data of the target production equipment and restructuring the real-time operating data to obtain the operating data base of the target production equipment includes: Collect real-time operating data of the target production equipment, including equipment status parameters, process execution indicators, and environmental interaction variables; Multi-source heterogeneous fusion is performed on the real-time running data to obtain a unified feature space for the real-time running data; Based on the unified feature space, tensor structure weaving is performed on the real-time running data to obtain the multidimensional feature tensor of the real-time running data. The multidimensional feature tensor is isomorphically mapped to the logical topology of the target production equipment to construct a topological association graph of the real-time running data; Based on the topological correlation graph, the multidimensional feature tensor is embedded in the manifold space to obtain the operating data basis of the target production equipment.

[0007] In a preferred embodiment, the step of constructing a three-dimensional geometric skeleton of the target production equipment based on its physical structure, and then synchronously binding the three-dimensional geometric skeleton with the operational data base to obtain an initial digital twin of the target production equipment, includes: The physical structure of the target production equipment is analyzed, and the geometric feature parameters and assembly hierarchy relationships of the target production equipment are extracted to obtain the skeleton topological constraints of the target production equipment. Based on the skeleton topological constraints, the geometric feature parameters are parametrically reconstructed to obtain the three-dimensional geometric skeleton of the target production equipment. Feature saliency detection is performed on the operational data base to obtain the spatiotemporal feature point cloud of the operational data base; The node coordinates of the three-dimensional geometric skeleton are spatially registered with the spatiotemporal feature point cloud to obtain the mapping relationship matrix between the three-dimensional geometric skeleton and the running data base. Based on the mapping matrix, the operational data base is semantically associated and bound with the three-dimensional geometric skeleton to obtain the initial digital twin of the target production equipment.

[0008] In a preferred embodiment, the step of parametrically reconstructing the geometric feature parameters of the target production equipment based on the skeleton topological constraints to obtain the three-dimensional geometric skeleton of the target production equipment includes: By performing geometric feature inverse extrapolation on the skeleton topological constraints, a parameterized representation basis for the skeleton topological constraints is obtained; Based on the parameterized expression basis, the geometric feature parameters are linearly combined using basis vectors to obtain the surface control points of the geometric feature parameters; The surface control points are topologically ordered and meshed to obtain parameterized surface patches of the surface control points. Based on the skeleton topological constraints, the parametric surface patches are hierarchically stitched together to obtain the three-dimensional geometric skeleton of the parametric surface patches.

[0009] In a preferred embodiment, the step of anchoring the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain an associated digital twin of the target production equipment includes: Fault mode analysis is performed on the historical fault records of the target production equipment to obtain the fault feature vector of the historical fault records; Extract the running parameter nodes from the initial digital twin, and perform parameter feature quantization on the running parameter nodes to obtain the motion feature vector of the running parameter nodes; The fault feature vector and the motion feature vector are spatially correlated and mapped to obtain the correlation confidence matrix between the fault feature vector and the motion feature vector. Based on the association confidence matrix, feature anchoring labels are applied to the operating parameter nodes and the historical fault records to obtain the fault association labels of the operating parameter nodes; Based on the fault association marker, the initial digital twin is updated in terms of topology to obtain the associated digital twin of the target production equipment.

[0010] In a preferred embodiment, the step of performing feature anchoring and labeling on the operating parameter nodes and the historical fault records based on the association confidence matrix to obtain the fault association label of the operating parameter nodes includes: Based on the correlation confidence matrix, the motion feature vector of the operating parameter node and the fault feature vector of the historical fault record are nonlinearly weighted and fused to obtain the enhanced confidence matrix of the correlation confidence matrix. The calculation formula of the enhanced confidence matrix is ​​as follows: ; in, In the enhanced confidence matrix, the first... Line number The element values ​​of the column, In the association confidence matrix, the first... Line number The element values ​​of the column, Indicates the first The motion feature vector of each running parameter node, Indicates the first The fault feature vector of each historical fault record, This indicates the preset scaling factor. The Euclidean norm of a vector; Based on the fusion confidence level, the candidate fault records of the operating parameter node are filtered by confidence level to obtain a high-confidence fault candidate set of the operating parameter node; Based on the high-confidence fault candidate set, a posterior attribution determination is performed on the operating parameter nodes to obtain the fault association markers of the operating parameter nodes.

[0011] In a preferred embodiment, the step of performing real-time anomaly threshold detection on the fault parameter nodes in the associated digital twin, and performing intelligent tracing and deduction on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment, includes: Transient parameter capture is performed on the fault parameter nodes in the associated digital twin to obtain the time series of the operating status of the fault parameter nodes; Anomaly drift detection is performed on the operating state time series to obtain the abnormal deviation trigger signal of the fault parameter node; Based on the abnormal deviation trigger signal, causal link tracing is performed on the associated digital twin to obtain the root cause node location of the associated digital twin; Based on the root cause node location, a fault propagation topology inversion is performed on the associated digital twin to obtain the fault propagation chain of the associated digital twin; By fusing and aggregating the root cause node location with the fault propagation chain, a preliminary diagnostic report for the target production equipment is obtained.

[0012] In a preferred embodiment, the step of hierarchically coupling and rendering the preliminary diagnostic report with the associated digital twin to obtain a visual operation and maintenance interface for the target production equipment includes: The preliminary diagnostic report is hierarchically analyzed to obtain a diagnostic information hierarchy tree of the preliminary diagnostic report; Based on the diagnostic information hierarchy tree, the geometric nodes of the associated digital twin are targeted and located to obtain the diagnostic anchoring nodes of the associated digital twin; Based on the diagnostic anchor node, the associated digital twin is visually mapped and encoded to obtain an enhanced visualization layer of the associated digital twin; The enhanced visualization layer and the basic geometric view of the associated digital twin are fused and rendered in a multimodal manner to obtain the visualized operation and maintenance interface of the target production equipment.

[0013] In a preferred embodiment, the step of compiling the interactive operation behavior of the visualized operation and maintenance interface into a visualized operation chain to obtain precise control and maintenance instructions for the target production equipment includes: The interactive operation behavior of the visual operation and maintenance interface is transcribed in a structured manner to obtain the operation event sequence of the interactive operation behavior; The semantic intent of the operation event sequence is deconstructed to obtain the operation intent vector of the operation event sequence; Based on the operation intent vector, the control parameter space of the target production equipment is mapped with instructions to obtain the operation primitive sequence of the operation event sequence; The operation primitive sequence is iteratively optimized and adapted to obtain precise control and maintenance instructions for the target production equipment.

[0014] To address the above problems, the present invention also provides a production equipment visualization operation and maintenance system based on digital twins, the system comprising: The data reconstruction module is used to collect real-time operating data of the target production equipment and perform structured reconstruction of the real-time operating data to obtain the operating data base of the target production equipment. The geometric binding module is used to construct a three-dimensional geometric skeleton of the target production equipment based on the physical structure of the target production equipment, and to bind the three-dimensional geometric skeleton to the running data base in a virtual-real synchronous manner to obtain an initial digital twin of the target production equipment. The fault anchoring module is used to anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment. The anomaly tracing module is used to perform real-time anomaly threshold detection on the fault parameter nodes in the associated digital twin, and to perform intelligent tracing and deduction on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment. The coupled rendering module is used to perform hierarchical coupled rendering of the preliminary diagnostic report and the associated digital twin to obtain a visual operation and maintenance interface for the target production equipment. The operation compilation module is used to compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain precise control and maintenance instructions for the target production equipment.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This technology constructs a data base with topological correlation characteristics by performing structured recombination operations such as multi-source heterogeneous fusion and tensor structured weaving on real-time operating data of production equipment. Combined with the physical structure of the equipment, it completes the parametric construction of a three-dimensional geometric skeleton. Through spatial coordinate registration and semantic association binding, it achieves synchronous fusion of virtual and real data, creating a digital twin that accurately maps the physical state and operating data of the equipment. Simultaneously, relying on a fault feature anchoring mechanism, it accurately associates the operating parameter nodes of the digital twin with historical fault records, enabling the digital twin to have fault feature perception capabilities. This significantly improves the authenticity, relevance, and relevance of the digital mapping of equipment operating status, providing accurate and comprehensive digital model support for equipment operation and maintenance.

[0016] 2. This technology performs real-time anomaly threshold detection on fault parameter nodes of the associated digital twin, and combines causal link tracing and fault propagation topology inversion to achieve intelligent fault tracing and deduction. This enables rapid and accurate location of the root cause of the fault and analysis of the fault propagation path, improving the efficiency and accuracy of equipment fault diagnosis. Through hierarchical coupled rendering, the diagnostic report is deeply integrated with the associated digital twin, creating a highly efficient and visualized operation and maintenance interface. The visualized operation chain then compiles the interface interaction operations into precise equipment control and maintenance commands, achieving integrated, visualized, and intelligent operation from fault diagnosis to operation and maintenance control. This significantly improves the overall efficiency of visualized operation and maintenance of production equipment, while making equipment operation and maintenance control more precise and timely. Attached Figure Description

[0017] Figure 1 A flowchart illustrating a digital twin-based visual operation and maintenance method for production equipment, as provided in an embodiment of the present invention; Figure 2 A functional module diagram of a production equipment visualization operation and maintenance system based on digital twins is provided in one embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for visualizing the operation and maintenance of production equipment based on digital twins. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, this method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cluster of cloud servers. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a digital twin-based visual operation and maintenance method for production equipment according to an embodiment of the present invention. In this embodiment, the digital twin-based visual operation and maintenance method for production equipment includes: P1. Collect real-time operating data of the target production equipment and perform structured reorganization on the real-time operating data to obtain the operating data base of the target production equipment; In this embodiment of the invention, the step of collecting real-time operating data of the target production equipment and structurally reorganizing the real-time operating data to obtain the operating data base of the target production equipment includes: Collect real-time operating data of the target production equipment, including equipment status parameters, process execution indicators, and environmental interaction variables; Multi-source heterogeneous fusion is performed on the real-time running data to obtain a unified feature space for the real-time running data; Based on the unified feature space, tensor structure weaving is performed on the real-time running data to obtain the multidimensional feature tensor of the real-time running data. The multidimensional feature tensor is isomorphically mapped to the logical topology of the target production equipment to construct a topological association graph of the real-time running data; Based on the topological correlation graph, the multidimensional feature tensor is embedded in the manifold space to obtain the operating data basis of the target production equipment.

[0021] For the equipment status parameters of the target production equipment, the physical operating status is continuously and in real time collected through dedicated sensing and monitoring devices mounted on each core component of the equipment. For the process execution indicators, the production process control system synchronously and in real time collects relevant indicators such as the execution accuracy of the equipment production process, the achievement of process parameters, and the production cycle. For environmental interaction variables, environmental monitoring terminals deployed at the equipment operation site continuously and in real time collect environmental data such as ambient temperature, humidity, air pressure, and dust concentration. The collected equipment status parameters, process execution indicators, and environmental interaction variables are fully integrated to form the real-time operating data of the target production equipment.

[0022] Data format standardization processing is carried out on real-time operational data collected from different sources and with different data structures. Unstructured environmental state description data, semi-structured process statistics data, and structured equipment sensor data are all converted into a unified and standardized structured data format. Then, core features are extracted from the standardized data to extract key feature information that can intuitively reflect the equipment operating status. Subsequently, all extracted key feature information is systematically classified and integrated according to the data attribute categories to build a unified feature space that can comprehensively cover all core features of real-time operational data.

[0023] Based on the unified feature space that has been constructed, the core feature information within the space is clearly divided according to the attribute dimensions of the data, resulting in different feature data dimensions such as equipment dimension, process dimension, and environment dimension. Then, the feature information under each dimension is systematically woven and integrated according to the chronological order of time and the spatial relationship, transforming the originally scattered feature information into tensor data with multi-dimensional attributes and a structured organizational form, thus forming a multi-dimensional feature tensor of real-time running data.

[0024] First, comprehensively analyze the logical topology of the target production equipment, clearly identifying the inherent logical relationships and hierarchical structures among the various components, production processes, and environmental influencing factors. Then, map and match the feature data of each dimension in the multidimensional feature tensor with the corresponding nodes and relationships in the equipment's logical topology. Based on the mapping and matching results, systematically construct the relationships between various types of data in the form of a graph. The nodes in the graph correspond to various data units related to equipment, processes, and environment, and the connections between nodes correspond to the logical relationships between data units, thereby constructing a topological relationship graph of real-time operating data.

[0025] Based on the complete topological relationships between real-time operational data presented by the topological relationship map, all data points in the multidimensional feature tensor are projected into the manifold space. Within the manifold space, the original topological relationship features and multidimensional attribute features of each data point are fully preserved. At the same time, the spatial dimension of the dataset after being projected into the manifold space is optimized and integrated, and redundant information in the dataset that does not affect the reflection of the equipment's operating status is removed. The core data set that can accurately and comprehensively reflect the actual operating status of the target production equipment is retained, and finally, the operational data base of the target production equipment is formed.

[0026] The beneficial effects are that by classifying and accurately collecting real-time operating data of the target production equipment and reorganizing it in a multi-step structured manner, standardized processing and deep integration of multi-source heterogeneous equipment, process and environmental data are achieved. The constructed operating data base not only retains the multi-dimensional attribute characteristics of the data, but also has clear topological correlation characteristics. It can accurately and comprehensively reflect the actual operating status of the target production equipment, and provides standardized, structured and highly correlated core data support for the subsequent virtual-real synchronous binding of the three-dimensional geometric skeleton and the operating data base. It ensures the accuracy and effectiveness of the subsequent digital twin construction from the data source.

[0027] P2. Based on the physical structure of the target production equipment, construct a three-dimensional geometric skeleton of the target production equipment, and bind the three-dimensional geometric skeleton to the operating data base in a virtual-real synchronous manner to obtain the initial digital twin of the target production equipment; In this embodiment of the invention, the step of constructing a three-dimensional geometric skeleton of the target production equipment based on its physical structure, and then binding the three-dimensional geometric skeleton to the operating data base in a virtual-real synchronization to obtain an initial digital twin of the target production equipment, includes: The physical structure of the target production equipment is analyzed, and the geometric feature parameters and assembly hierarchy relationships of the target production equipment are extracted to obtain the skeleton topological constraints of the target production equipment. Based on the skeleton topological constraints, the geometric feature parameters are parametrically reconstructed to obtain the three-dimensional geometric skeleton of the target production equipment. Feature saliency detection is performed on the operational data base to obtain the spatiotemporal feature point cloud of the operational data base; The node coordinates of the three-dimensional geometric skeleton are spatially registered with the spatiotemporal feature point cloud to obtain the mapping relationship matrix between the three-dimensional geometric skeleton and the running data base. Based on the mapping matrix, the operational data base is semantically associated and bound with the three-dimensional geometric skeleton to obtain the initial digital twin of the target production equipment.

[0028] The step of parametrically reconstructing the geometric feature parameters of the target production equipment based on the skeleton topological constraints to obtain the three-dimensional geometric skeleton of the target production equipment includes: By performing geometric feature inverse extrapolation on the skeleton topological constraints, a parameterized representation basis for the skeleton topological constraints is obtained; Based on the parameterized expression basis, the geometric feature parameters are linearly combined using basis vectors to obtain the surface control points of the geometric feature parameters; The surface control points are topologically ordered and meshed to obtain parameterized surface patches of the surface control points. Based on the skeleton topological constraints, the parametric surface patches are hierarchically stitched together to obtain the three-dimensional geometric skeleton of the parametric surface patches.

[0029] A comprehensive structural scan and disassembly analysis of the physical entity of the target production equipment is conducted to clarify the geometric attributes of each component, such as its outline, dimensions, and surface shape. Geometric feature parameters that fully reflect the geometric characteristics of each component are extracted. At the same time, the assembly sequence, connection method, and hierarchical relationship of the equipment from the base to each functional module and then to the top-level execution component are systematically sorted out. The extracted geometric feature parameters are integrated with the sorted assembly hierarchy relationship and the mutual constraint rules are clarified to obtain the skeleton topology constraints of the target production equipment.

[0030] Based on the assembly hierarchy and geometric feature parameter association constraint rules clearly defined in the skeleton topology constraints, the topology of the overall equipment is reverse-engineered to the local geometric features of each component. Basic feature elements that can characterize the surface morphology construction rules of the overall equipment and each component are extracted. These basic feature elements are then integrated in an orderly manner according to the topological association relationship to form a standard basis that can guide the geometric feature parameters to carry out surface reconstruction, thus obtaining the parameterized expression basis of the skeleton topology constraints.

[0031] The geometric feature parameters of each component of the target production equipment are matched one by one with the basic feature primitives in the parameterized expression basis. According to the surface construction rules of the primitives, the corresponding geometric feature parameters are combined and extended in an orderly manner, so that the originally discrete geometric feature parameters are transformed into feature point information that can support the surface forming of the equipment, and the surface control points of the geometric feature parameters are obtained.

[0032] Based on the equipment assembly hierarchy and spatial topological relationships between components as defined in the skeleton topology constraints, all surface control points are arranged in an orderly manner in space. Adjacent surface control points with direct topological relationships are connected in space to construct a continuous network structure, allowing discrete surface control points to form planar basic units with complete spatial topology structures, thus obtaining parameterized surface patches of surface control points.

[0033] Following the assembly hierarchy of the target production equipment from the bottom basic components to the top functional components, the parametric surface patches corresponding to each component are precisely spliced ​​and merged according to the topological association rules in the skeleton topological constraints. During the splicing process, the connection position and splicing angle of each surface patch are kept highly consistent with the actual assembly state of the physical equipment. All independent parametric surface patches are integrated into a spatial structure that can completely represent the overall physical form of the target production equipment, thus obtaining the three-dimensional geometric skeleton of the target production equipment.

[0034] A full-dimensional feature analysis is performed on all data contained in the operational data base to identify the core feature data that can accurately reflect the operating status of different parts of the equipment and has the highest correlation with the operation of the equipment's physical structure. These core feature data are then processed into points according to the time change sequence and the corresponding spatial location of the equipment, forming a point cloud set composed of a series of feature points with spatiotemporal attributes, thus obtaining the spatiotemporal feature point cloud of the operational data base.

[0035] All structural nodes of the 3D geometric skeleton are calibrated in a unified spatial coordinate system to clarify the precise 3D spatial coordinate information of each structural node. At the same time, all feature points in the spatiotemporal feature point cloud are calibrated in the same spatial coordinate system. The spatial coordinates of the two are matched and calibrated one by one to achieve a precise spatial position correspondence between the structural nodes of the 3D geometric skeleton and the feature points of the spatiotemporal feature point cloud in the same spatial coordinate system. Based on the coordinate correspondence results, a matrix-form data that clearly reflects the spatial correspondence between the two is constructed to obtain the mapping relationship matrix between the 3D geometric skeleton and the running data base.

[0036] Based on the mapping relationship matrix, various types of data in the operational data base are semantically associated and matched with the corresponding structural nodes in the 3D geometric skeleton. This ensures that each structural node in the 3D geometric skeleton can be uniquely associated with relevant data reflecting its actual operating status. At the same time, a dynamic association mechanism is established to synchronize the real-time updates of the operational data base with the synchronous changes in the state of the 3D geometric skeleton. This allows the 3D geometric skeleton to reflect the actual operating status of the equipment in real time as the values ​​of the operational data base change. The 3D geometric skeleton, which incorporates the operational data base, is then integrated as a whole to obtain the initial digital twin of the target production equipment.

[0037] The beneficial effects are that by accurately analyzing the physical structure of the target production equipment and reconstructing the parametric surface, the constructed three-dimensional geometric skeleton can completely and accurately represent the actual physical form of the equipment. Combined with feature saliency detection and spatial coordinate registration, the three-dimensional geometric skeleton and the operating data base are accurately and synchronously bound to the virtual and real. The resulting initial digital twin not only has the same geometric form as the physical equipment, but can also map the actual operating status of the equipment in real time through data association. This achieves a deep integration of the physical form of the equipment and the operating data, providing an accurate, complete and dynamic digital twin foundation for subsequent fault feature anchoring and intelligent source tracing and inference, and ensuring the efficiency and accuracy of subsequent operation and maintenance.

[0038] P3. Anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment; In this embodiment of the invention, the step of anchoring the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain an associated digital twin of the target production equipment includes: Fault mode analysis is performed on the historical fault records of the target production equipment to obtain the fault feature vector of the historical fault records; Extract the running parameter nodes from the initial digital twin, and perform parameter feature quantization on the running parameter nodes to obtain the motion feature vector of the running parameter nodes; The fault feature vector and the motion feature vector are spatially correlated and mapped to obtain the correlation confidence matrix between the fault feature vector and the motion feature vector. Based on the association confidence matrix, feature anchoring labels are applied to the operating parameter nodes and the historical fault records to obtain the fault association labels of the operating parameter nodes; Based on the fault association marker, the initial digital twin is updated in terms of topology to obtain the associated digital twin of the target production equipment.

[0039] The step of performing feature anchoring and labeling on the operating parameter nodes and the historical fault records based on the association confidence matrix to obtain the fault association label of the operating parameter nodes includes: Based on the correlation confidence matrix, the motion feature vector of the operating parameter node and the fault feature vector of the historical fault record are nonlinearly weighted and fused to obtain the enhanced confidence matrix of the correlation confidence matrix. The calculation formula of the enhanced confidence matrix is ​​as follows: ; in, In the enhanced confidence matrix, the first... Line number The element values ​​of the column, In the association confidence matrix, the first... Line number The element values ​​of the column, Indicates the first The motion feature vector of each running parameter node, Indicates the first The fault feature vector of each historical fault record, This indicates the preset scaling factor. The Euclidean norm of a vector; Based on the fusion confidence level, the candidate fault records of the operating parameter node are filtered by confidence level to obtain a high-confidence fault candidate set of the operating parameter node; Based on the high-confidence fault candidate set, a posterior attribution determination is performed on the operating parameter nodes to obtain the fault association markers of the operating parameter nodes.

[0040] A systematic review of all historical fault records of the target production equipment is conducted. Core information such as the location of the fault, the specific manifestation of the fault, the environmental conditions that triggered the fault, and the equipment operating status at the time of the fault is extracted from each fault record. The extracted core information is then standardized and organized into features. The core fault features of each historical fault record are then visualized according to a unified feature dimension. The represented fault features are then transformed into vector data with fixed feature dimensions to obtain the fault feature vector of the historical fault record.

[0041] Iterate through all structural nodes of the initial digital twin, and select the operating parameter nodes that directly correspond to the equipment operating parameters and can reflect the real-time operating status of the equipment based on the actual logic of equipment operation. Clarify the equipment operating parameter type, parameter monitoring dimension, and correlation dimension of parameter changes for each operating parameter node. Perform comprehensive feature extraction and quantification on the parameter value characteristics, parameter change rules, and correlation attributes between parameters for each operating parameter node. Organize the quantified parameter features according to the feature dimension consistent with the fault feature vector, and transform the feature information of each operating parameter node into vector data with the same feature dimension to obtain the motion feature vector of the operating parameter node.

[0042] A unified feature analysis space is constructed, and all fault feature vectors and motion feature vectors are projected into this feature analysis space. The feature similarity, attribute correlation, and logical matching between each motion feature vector and each fault feature vector in the space are analyzed one by one. The correlation between each set of motion feature vectors and fault feature vectors is quantitatively represented. All quantitative representation results are arranged in an orderly manner according to the correspondence between motion feature vectors and fault feature vectors, with rows corresponding to motion feature vectors of operating parameter nodes and columns corresponding to fault feature vectors of historical fault records. The elements in the matrix directly represent the correlation between the two sets of vectors, thus obtaining the correlation confidence matrix between fault feature vectors and motion feature vectors.

[0043] Based on the initial correlation degree quantification results in the correlation confidence matrix, targeted weight adjustments are made by combining the spatial distance between motion feature vectors and fault feature vectors in a unified feature analysis space and the fit of feature matching. The weight ratio of matrix elements corresponding to vectors with high feature similarity and close spatial distance is increased, while the weight ratio of matrix elements corresponding to vectors with low feature similarity and far spatial distance is decreased. In this way, the nonlinear weighted fusion of motion feature vectors of operating parameter nodes and fault feature vectors of historical fault records is achieved. All correlation degree quantification results after weight adjustment are re-integrated according to the original matrix arrangement to obtain the enhanced confidence matrix of the correlation confidence matrix.

[0044] The enhanced confidence matrix Line number The element values ​​of the column are derived from the element values ​​at the corresponding positions in the association confidence matrix, and the first... The motion feature vector of the nth running parameter node, the nth The fault feature vector of the first historical fault record and the preset scaling factor are associated with the confidence matrix of the first... Line number The element value of the column represents the first The motion feature vector of the nth running parameter node and the nth The initial correlation between the fault feature vectors of each historical fault record is represented by the feature similarity, attribute correlation, and logical matching of the two types of vectors analyzed in the previous spatial correlation mapping step. The motion feature vector of each running parameter node originates from the extraction and quantification of the parameter value characteristics, parameter variation patterns, and inter-parameter correlation attributes of the corresponding running parameter nodes in the initial digital twin. It is the core vector data reflecting the real-time running status of the running parameter nodes. The fault feature vector of the first historical fault record is derived from the fault mode analysis of the historical fault records. It is vector data representing the core features of historical faults. The preset scaling factor is a fixed value used to adjust the weight ratio of vector feature matching. It can be preset and determined according to the operation and maintenance requirements of the target production equipment, the fault feature type, and the importance of the operating parameter nodes. After the above parameters are integrated, they are used to generate the first enhanced confidence matrix through corresponding calculation rules. Line number The element values ​​of the column.

[0045] This formula is used to optimize and adjust the element values ​​in the correlation confidence matrix that represent the correlation between operating parameter nodes and historical fault records. It calculates the... The motion feature vector of the nth running parameter node and the nth The matching degree of the fault feature vectors of historical fault records in the feature space is used to adjust the initial correlation degree value nonlinearly. This accurately corrects the correlation confidence between the two types of vectors, so that the adjusted correlation confidence can better reflect the real correlation between the running parameter nodes and historical fault records in the actual scenario. This provides a more accurate confidence basis for the subsequent screening of high-confidence fault candidate sets, ensuring the accuracy and reliability of the fault feature anchoring step.

[0046] The formula exhibits a trend of adjustment as the matching degree between the motion feature vector and the fault feature vector changes. When the feature similarity between the motion feature vector and the fault feature vector is higher, the spatial distance is closer, and the matching degree value is larger, the formula calculation result shows a greater improvement in the initial association confidence value compared to the initial association confidence value, demonstrating the strengthening effect on high-matching association relationships. When the feature similarity between the motion feature vector and the fault feature vector is lower, the spatial distance is farther, and the matching degree value is smaller, the adjustment range of the formula calculation result compared to the initial association confidence value is smaller, and even when the matching degree is extremely low, the initial association confidence value remains unchanged, demonstrating the weakening effect on low-matching association relationships. At the same time, the preset scaling factor can further adjust the magnitude of the adjustment range to adapt to the fault association feature analysis needs of different target production equipment.

[0047] A unified confidence level screening criterion is established, with the element values ​​in the enhanced confidence level matrix as the core judgment basis. The association confidence level of all historical fault records corresponding to each operating parameter node is checked and judged one by one, and historical fault records whose confidence level values ​​meet the preset screening criteria are selected. All historical fault records corresponding to each operating parameter node that meet the screening criteria are centrally integrated to build a dedicated fault record set for each operating parameter node, thus obtaining a high-confidence fault candidate set for the operating parameter node.

[0048] For each operating parameter node, a secondary feature matching verification is performed on all historical fault records in the high-confidence fault candidate set. Combining the actual operating logic of the target production equipment and the associated operating mechanism of each component, the causal relationship and mutual influence relationship between each fault record in the candidate set and the corresponding operating parameter node are analyzed. The final effective association relationship between each operating parameter node and each historical fault record in the candidate set is clarified. According to the unified marking rules, a unique identifier containing the associated fault record type, core fault characteristics, and fault association degree is added to each operating parameter node to obtain the fault association mark of the operating parameter node.

[0049] The fault association tags of all operating parameter nodes are fully integrated into the topology of the initial digital twin. Within the topology of the digital twin, a module for storing and displaying fault association information is added to each operating parameter node, allowing each operating parameter node to be directly associated with the corresponding historical fault record information. At the same time, the topology association rules of the initial digital twin are updated according to the fault association tags, making the fault association features an inherent component of the digital twin topology. This completes the full-dimensional update and optimization of the initial digital twin topology. By integrating the updated topology information, operating parameter information, and fault association information, the associated digital twin of the target production equipment is obtained.

[0050] The beneficial effects are as follows: by standardizing and vectorizing historical fault records and operating parameter nodes, a unified dimensional analysis of fault features and operating parameter features is achieved. Spatial correlation mapping and nonlinear weighted fusion improve the accuracy of the correlation and matching between operating parameter nodes and historical fault records. Then, through confidence screening and posterior attribution determination, accurate fault association labels are obtained. Finally, a digital twin with updated topology structure is completed, enabling precise feature anchoring and information binding between operating parameter nodes and historical fault records. This gives the digital twin the core capability of fault feature association, which can quickly match and associate real-time equipment operating parameters with historical faults. It provides a digital twin foundation with fault association features for subsequent anomaly detection and intelligent source tracing of fault parameter nodes, greatly improving the pertinence and execution efficiency of subsequent fault diagnosis.

[0051] P4. Real-time abnormal threshold detection is performed on the fault parameter nodes in the associated digital twin, and intelligent source tracing and deduction are performed on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment. In this embodiment of the invention, the step of performing real-time anomaly threshold detection on the fault parameter nodes in the associated digital twin, and performing intelligent tracing and deduction on the associated digital twin based on the detected trigger signal to obtain a preliminary diagnostic report for the target production equipment, includes: Transient parameter capture is performed on the fault parameter nodes in the associated digital twin to obtain the time series of the operating status of the fault parameter nodes; Anomaly drift detection is performed on the operating state time series to obtain the abnormal deviation trigger signal of the fault parameter node; Based on the abnormal deviation trigger signal, causal link tracing is performed on the associated digital twin to obtain the root cause node location of the associated digital twin; Based on the root cause node location, a fault propagation topology inversion is performed on the associated digital twin to obtain the fault propagation chain of the associated digital twin; By fusing and aggregating the root cause node location with the fault propagation chain, a preliminary diagnostic report for the target production equipment is obtained.

[0052] Traverse all fault parameter nodes marked with fault association tags in the associated digital twin. For the equipment operation parameters corresponding to each fault parameter node, continuously collect the real-time operation status parameters of the node in chronological order. Integrate the parameter information collected at each time node in chronological order to form a continuous data sequence that can completely reflect the changes in the operation status of the fault parameter node at different time points, and obtain the operation status time series of the fault parameter node.

[0053] Using the historical normal operating status data of the fault parameter node as a benchmark, the parameter information of each time node in the operating status time series is compared point by point to identify the degree of deviation and trend of the parameter relative to the normal operating status. When a parameter is detected to have a continuous drift from the normal operating range, a trigger signal for the corresponding fault parameter node is generated to obtain the abnormal deviation trigger signal of the fault parameter node.

[0054] Upon receiving an abnormal deviation trigger signal, starting from the fault parameter node corresponding to the trigger signal, the system traces the causal relationship between each node in the link along the established topological link in the associated digital twin. It then examines the impact of each upstream node on the current fault node's operating status, identifies the transmission path and impact logic of the abnormal state, and finally locates the initial source node that caused the abnormal drift, thus obtaining the root cause node location in the associated digital twin.

[0055] Starting from the initial source node determined by the root cause node location, the transmission process of the fault state from the root cause node to each downstream node is deduced forward along the topological association links already constructed in the associated digital twin. The order of fault impact, transmission path and degree of impact between each node are clarified. The entire fault transmission path, logic and impact relationship are systematically sorted out and integrated to form a complete fault state transmission link, and the fault propagation chain of the associated digital twin is obtained.

[0056] The fault source information identified by the root cause node location is comprehensively integrated with the fault propagation chain to sort out the fault transmission path and impact logic, combined with the abnormal state information of the corresponding fault parameter nodes and historical fault association information. All information is systematically collected and organized according to a unified report format to form a complete diagnostic content including the fault root cause, fault propagation path and abnormal equipment status, and a preliminary diagnostic report of the target production equipment is obtained.

[0057] The beneficial effects are that by capturing transient parameters and detecting abnormal drift of fault parameter nodes in the associated digital twin, real-time and accurate identification of abnormal equipment states is achieved. Relying on causal link tracing and fault propagation topology inversion, the root cause of the fault is accurately located and the fault propagation path is fully sorted out. The resulting preliminary diagnostic report comprehensively covers the root cause of the fault, propagation path and abnormal state information, providing accurate fault diagnosis basis for the subsequent hierarchical coupling rendering of the visual operation and maintenance interface and the generation of precise control instructions. This greatly improves the efficiency and accuracy of fault diagnosis of production equipment and ensures the timeliness and effectiveness of equipment operation and maintenance.

[0058] P5. Perform hierarchical coupling rendering of the preliminary diagnostic report and the associated digital twin to obtain the visual operation and maintenance interface of the target production equipment; In this embodiment of the invention, the step of hierarchically coupling and rendering the preliminary diagnostic report with the associated digital twin to obtain the visual operation and maintenance interface of the target production equipment includes: The preliminary diagnostic report is hierarchically analyzed to obtain a diagnostic information hierarchy tree of the preliminary diagnostic report; Based on the diagnostic information hierarchy tree, the geometric nodes of the associated digital twin are targeted and located to obtain the diagnostic anchoring nodes of the associated digital twin; Based on the diagnostic anchor node, the associated digital twin is visually mapped and encoded to obtain an enhanced visualization layer of the associated digital twin; The enhanced visualization layer and the basic geometric view of the associated digital twin are fused and rendered in a multimodal manner to obtain the visualized operation and maintenance interface of the target production equipment.

[0059] The preliminary diagnostic report is disassembled to extract root cause information, fault propagation path information, equipment abnormal status information, and historical fault association information. The information is then hierarchically divided according to its logical hierarchy and impact priority. The overall fault diagnosis conclusion is the top-level node, the root cause node information is the second-level node, the information of each link in the fault propagation chain is the middle-level node, and the abnormal status information and historical fault association information of each fault parameter node are the bottom-level nodes. All nodes are then connected in an orderly manner according to their hierarchical hierarchy to form a complete and logically clear diagnostic information hierarchy tree.

[0060] Traverse all hierarchical nodes in the diagnostic information hierarchy tree, extract the physical structure location information and operating parameter node information of the target production equipment corresponding to each node, match the extracted location information and operating parameter information with each geometric node of the three-dimensional geometric skeleton in the associated digital twin, accurately locate all geometric nodes directly related to the diagnostic information, and use these located geometric nodes as dedicated nodes to carry the diagnostic information to obtain the diagnostic anchoring nodes of the associated digital twin.

[0061] For each diagnostic anchor node, a unique visual identification rule is assigned to each diagnostic anchor node based on the node level and information type in the diagnostic information hierarchy tree. The visual features of the diagnostic anchor nodes are enhanced according to the rules, including the display style, highlighting method, and information labeling content. All the visually enhanced diagnostic anchor nodes are integrated according to the topology of the associated digital twin to form a unique visualization layer superimposed on the basic geometric view, resulting in an enhanced visualization layer of the associated digital twin.

[0062] The enhanced visualization layer is precisely overlaid on the basic geometric view of the associated digital twin according to the spatial coordinate correspondence, ensuring that each diagnostic anchor node in the enhanced visualization layer completely coincides with the corresponding geometric node in the basic geometric view. The visual effects of the overlaid view are optimized to ensure that the diagnostic information is clearly and prominently displayed without affecting the normal display of the basic geometric view. At the same time, an interactive response mechanism is added to the view, allowing maintenance personnel to view diagnostic information at different levels through the operation view. Finally, an interface integrating equipment 3D geometric display, fault diagnosis information visualization, and interactive operation is formed, resulting in a visualized maintenance interface for the target production equipment.

[0063] The beneficial effects are as follows: by hierarchically sorting out the preliminary diagnostic report, a logically clear hierarchical tree of diagnostic information is constructed; by relying on targeted matching and positioning, the accurate correspondence between diagnostic information and the geometric nodes of the associated digital twin is achieved; the enhanced visualization layer generated by visual mapping and encoding realizes the intuitive visual enhancement of fault diagnosis information; and by using multimodal fusion rendering, a visual operation and maintenance interface integrating equipment 3D display, fault information visualization, and interactive operation is created, which greatly improves the efficiency of fault diagnosis information transmission and the intuitiveness of display, provides operation and maintenance personnel with a clear and comprehensive carrier for displaying equipment operating status and fault information, and ensures the convenience and accuracy of operation and maintenance operations.

[0064] P6. Compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain the precise control and operation and maintenance instructions for the target production equipment.

[0065] In this embodiment of the invention, the step of compiling the interactive operation behavior of the visualized operation and maintenance interface into a visualized operation chain to obtain precise control and maintenance instructions for the target production equipment includes: The interactive operation behavior of the visual operation and maintenance interface is transcribed in a structured manner to obtain the operation event sequence of the interactive operation behavior; The semantic intent of the operation event sequence is deconstructed to obtain the operation intent vector of the operation event sequence; Based on the operation intent vector, the control parameter space of the target production equipment is mapped with instructions to obtain the operation primitive sequence of the operation event sequence; The operation primitive sequence is iteratively optimized and adapted to obtain precise control and maintenance instructions for the target production equipment.

[0066] All interactive operations generated on the visual operation and maintenance interface are fully captured, covering various operations such as clicked diagnostic anchor nodes, selected function options, triggered operation buttons, and input of control-related information. Each captured interactive operation is labeled with standardized information such as operation type, operation object, operation action, and operation occurrence time. According to the actual occurrence time of the operation, all standardized interactive operations are arranged in an orderly manner to form a continuous operation sequence with a unified structure and clear operation information, thus obtaining the operation event sequence of interactive operation behavior.

[0067] Traverse each operation event in the operation event sequence, extract the core operation information and the operation target object contained in each operation event, combine the functional design logic of the visual operation and maintenance interface and the actual business logic of the target production equipment operation and maintenance, analyze the specific operation purpose corresponding to each individual operation event, and then integrate the operation purposes of all operation events in the entire operation event sequence to sort out the overall operation and maintenance operation intent corresponding to the operation event sequence. According to the preset intent feature dimensions that cover the entire equipment operation and maintenance scenario, the sorted overall operation intent is comprehensively characterized, and the characterized overall operation intent information is transformed into vector data with clear dimensional features to obtain the operation intent vector of the operation event sequence.

[0068] A comprehensive analysis of the control parameter space of the target production equipment is conducted to identify all types of adjustable parameters, their control ranges, control methods, and the basic operation commands corresponding to various control behaviors. The features of each dimension in the operation intent vector are matched and logically correlated with the basic operation commands in the control parameter space. Based on the matching and correlation results, the basic operation commands that perfectly match the overall operation intent are precisely extracted. Then, according to the inherent execution logic of the operation intent and the conventional operation sequence of the target production equipment, all extracted basic operation commands are arranged in an orderly manner to form a continuous command sequence composed of basic operation commands, resulting in a sequence of operation primitives for the operation event sequence.

[0069] The system retrieves real-time operating status data, current fault diagnosis information, and the equipment's hardware operating logic and process execution constraints of the target production equipment. Based on this, it conducts a comprehensive rationality verification of each basic operation instruction in the operation primitive sequence, eliminating basic operation instructions that are inconsistent with the current operating status of the equipment or violate the equipment's operating logic. The execution details and execution order of the remaining basic operation instructions are adjusted according to the actual control needs of the equipment. The adjusted basic operation instructions are then systematically integrated, and the connection links between instructions are optimized according to the process requirements of on-site operation of the target production equipment. This ensures that the execution logic of the entire instruction sequence is highly consistent with the actual control needs of the equipment, ultimately forming continuous control instructions with direct executableness and precise control, resulting in precise control and maintenance instructions for the target production equipment.

[0070] The beneficial effects are as follows: By structuring and transcribing the interactive operations of the visual operation and maintenance interface, the fragmented operation behaviors are standardized and integrated. Through semantic intent deconstruction, the overall operation intent of the operation and maintenance personnel can be accurately mined. Relying on the instruction mapping of the equipment control parameter space, the operation intent is transformed into a basic sequence of operation primitives. Then, combined with the real-time operating status of the equipment, fault diagnosis information, and operating logic, iterative optimization and adaptation are completed. The generated precise control and maintenance instructions can accurately match the operation intent of the operation and maintenance personnel with the actual control needs of the equipment. This achieves accurate and efficient transformation from the visual interactive operation of the interface to the actual control instructions of the equipment, ensuring the executability and adaptability of the control instructions. It allows the visual operation and maintenance of production equipment to form a complete closed loop from fault diagnosis, interface interaction to precise control, which greatly improves the accuracy, timeliness and overall execution efficiency of production equipment operation and maintenance control.

[0071] like Figure 2 The diagram shown is a functional block diagram of a production equipment visualization operation and maintenance system based on digital twins provided in an embodiment of the present invention.

[0072] The digital twin-based production equipment visualization operation and maintenance system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the digital twin-based production equipment visualization operation and maintenance system 100 may include a data reconstruction module 101, a geometric binding module 102, a fault anchoring module 103, an anomaly tracing module 104, a coupled rendering module 105, and an operation compilation module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0073] In this embodiment, the functions of each module / unit are as follows: The data reconstruction module 101 is used to collect real-time operating data of the target production equipment and perform structured reconstruction of the real-time operating data to obtain the operating data base of the target production equipment. The geometric binding module 102 is used to construct a three-dimensional geometric skeleton of the target production equipment according to the physical structure of the target production equipment, and to bind the three-dimensional geometric skeleton to the running data base in a virtual-real synchronous manner to obtain an initial digital twin of the target production equipment. The fault anchoring module 103 is used to anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment. The anomaly tracing module 104 is used to perform real-time anomaly threshold detection on the fault parameter nodes in the associated digital twin, and to perform intelligent tracing and deduction on the associated digital twin based on the detected trigger signal to obtain a preliminary diagnostic report of the target production equipment. The coupling rendering module 105 is used to perform hierarchical coupling rendering of the preliminary diagnostic report and the associated digital twin to obtain a visual operation and maintenance interface for the target production equipment. The operation compilation module 106 is used to compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain precise control and maintenance instructions for the target production equipment.

[0074] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0075] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0076] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0077] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0078] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for visualized operation and maintenance of production equipment based on digital twins, characterized in that, The method includes: P1. Collect real-time operating data of the target production equipment and perform structured reorganization on the real-time operating data to obtain the operating data base of the target production equipment; P2. Based on the physical structure of the target production equipment, construct a three-dimensional geometric skeleton of the target production equipment, and bind the three-dimensional geometric skeleton to the operating data base in a virtual-real synchronous manner to obtain the initial digital twin of the target production equipment; P3. Anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment; P4. Real-time abnormal threshold detection is performed on the fault parameter nodes in the associated digital twin, and intelligent source tracing and deduction are performed on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment. P5. Perform hierarchical coupling rendering of the preliminary diagnostic report and the associated digital twin to obtain the visual operation and maintenance interface of the target production equipment; P6. Compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain the precise control and operation and maintenance instructions for the target production equipment.

2. The method for visualized operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The process of collecting real-time operating data from the target production equipment and then reorganizing that real-time operating data into a structured data base to obtain the operating data of the target production equipment includes: Collect real-time operating data of the target production equipment, including equipment status parameters, process execution indicators, and environmental interaction variables; Multi-source heterogeneous fusion is performed on the real-time running data to obtain a unified feature space for the real-time running data; Based on the unified feature space, tensor structure weaving is performed on the real-time running data to obtain the multidimensional feature tensor of the real-time running data. The multidimensional feature tensor is isomorphically mapped to the logical topology of the target production equipment to construct a topological association graph of the real-time running data; Based on the topological correlation graph, the multidimensional feature tensor is embedded in the manifold space to obtain the operating data basis of the target production equipment.

3. The method for visualizing the operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The process of constructing a three-dimensional geometric skeleton of the target production equipment based on its physical structure, and then binding the three-dimensional geometric skeleton to the operational data base in a virtual-real synchronization to obtain an initial digital twin of the target production equipment includes: The physical structure of the target production equipment is analyzed, and the geometric feature parameters and assembly hierarchy relationships of the target production equipment are extracted to obtain the skeleton topological constraints of the target production equipment. Based on the skeleton topological constraints, the geometric feature parameters are parametrically reconstructed to obtain the three-dimensional geometric skeleton of the target production equipment. Feature saliency detection is performed on the operational data base to obtain the spatiotemporal feature point cloud of the operational data base; The node coordinates of the three-dimensional geometric skeleton are spatially registered with the spatiotemporal feature point cloud to obtain the mapping relationship matrix between the three-dimensional geometric skeleton and the running data base. Based on the mapping matrix, the operational data base is semantically associated and bound with the three-dimensional geometric skeleton to obtain the initial digital twin of the target production equipment.

4. The method for visualized operation and maintenance of production equipment based on digital twins as described in claim 3, characterized in that, The step of parametrically reconstructing the geometric feature parameters of the target production equipment based on the skeleton topological constraints to obtain the three-dimensional geometric skeleton of the target production equipment includes: By performing geometric feature inverse extrapolation on the skeleton topological constraints, a parameterized representation basis for the skeleton topological constraints is obtained; Based on the parameterized expression basis, the geometric feature parameters are linearly combined using basis vectors to obtain the surface control points of the geometric feature parameters; The surface control points are topologically ordered and meshed to obtain parameterized surface patches of the surface control points. Based on the skeleton topological constraints, the parametric surface patches are hierarchically stitched together to obtain the three-dimensional geometric skeleton of the parametric surface patches.

5. The method for visualizing the operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The step of anchoring the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment includes: Fault mode analysis is performed on the historical fault records of the target production equipment to obtain the fault feature vector of the historical fault records; Extract the running parameter nodes from the initial digital twin, and perform parameter feature quantization on the running parameter nodes to obtain the motion feature vector of the running parameter nodes; The fault feature vector and the motion feature vector are spatially correlated and mapped to obtain the correlation confidence matrix between the fault feature vector and the motion feature vector. Based on the association confidence matrix, feature anchoring labels are applied to the operating parameter nodes and the historical fault records to obtain the fault association labels of the operating parameter nodes; Based on the fault association marker, the initial digital twin is updated in terms of topology to obtain the associated digital twin of the target production equipment.

6. The method for visualized operation and maintenance of production equipment based on digital twins as described in claim 5, characterized in that, The step of performing feature anchoring and labeling on the operating parameter nodes and the historical fault records based on the association confidence matrix to obtain the fault association label of the operating parameter nodes includes: Based on the correlation confidence matrix, the motion feature vector of the operating parameter node and the fault feature vector of the historical fault record are nonlinearly weighted and fused to obtain the enhanced confidence matrix of the correlation confidence matrix. The calculation formula of the enhanced confidence matrix is ​​as follows: ； in, Indicates the first element in the enhanced confidence matrix. Line number The element values ​​of the column, In the association confidence matrix, the first... Line number The element values ​​of the column, Indicates the first The motion feature vector of each running parameter node, Indicates the first The fault feature vector of each historical fault record, This indicates the preset scaling factor. The Euclidean norm of a vector; Based on the fusion confidence level, the candidate fault records of the operating parameter node are filtered by confidence level to obtain a high-confidence fault candidate set of the operating parameter node; Based on the high-confidence fault candidate set, a posterior attribution determination is performed on the operating parameter nodes to obtain the fault association markers of the operating parameter nodes.

7. The method for visualized operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The process involves real-time anomaly threshold detection of fault parameter nodes in the associated digital twin, and intelligent tracing and deduction of the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report for the target production equipment, including: Transient parameter capture is performed on the fault parameter nodes in the associated digital twin to obtain the time series of the operating status of the fault parameter nodes; Anomaly drift detection is performed on the operating state time series to obtain the abnormal deviation trigger signal of the fault parameter node; Based on the abnormal deviation trigger signal, causal link tracing is performed on the associated digital twin to obtain the root cause node location of the associated digital twin; Based on the root cause node location, a fault propagation topology inversion is performed on the associated digital twin to obtain the fault propagation chain of the associated digital twin; By fusing and aggregating the root cause node location with the fault propagation chain, a preliminary diagnostic report for the target production equipment is obtained.

8. The method for visualizing the operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The step of hierarchically coupling and rendering the preliminary diagnostic report with the associated digital twin to obtain a visual operation and maintenance interface for the target production equipment includes: The preliminary diagnostic report is hierarchically analyzed to obtain a diagnostic information hierarchy tree of the preliminary diagnostic report; Based on the diagnostic information hierarchy tree, the geometric nodes of the associated digital twin are targeted and located to obtain the diagnostic anchoring nodes of the associated digital twin; Based on the diagnostic anchor node, the associated digital twin is visually mapped and encoded to obtain an enhanced visualization layer of the associated digital twin; The enhanced visualization layer and the basic geometric view of the associated digital twin are fused and rendered in a multimodal manner to obtain the visualized operation and maintenance interface of the target production equipment.

9. A method for visualized operation and maintenance of production equipment based on digital twins as described in claim 1, characterized in that, The step of compiling the interactive operation behavior of the visualized operation and maintenance interface into a visualized operation chain to obtain precise control and maintenance instructions for the target production equipment includes: The interactive operation behavior of the visual operation and maintenance interface is transcribed in a structured manner to obtain the operation event sequence of the interactive operation behavior; The semantic intent of the operation event sequence is deconstructed to obtain the operation intent vector of the operation event sequence; Based on the operation intent vector, the control parameter space of the target production equipment is mapped with instructions to obtain the operation primitive sequence of the operation event sequence; The operation primitive sequence is iteratively optimized and adapted to obtain precise control and maintenance instructions for the target production equipment.

10. A production equipment visualization operation and maintenance system based on digital twins, characterized in that, The system for implementing the digital twin-based visual operation and maintenance method for production equipment as described in claim 1 includes: The data reconstruction module is used to collect real-time operating data of the target production equipment and perform structured reconstruction of the real-time operating data to obtain the operating data base of the target production equipment. The geometric binding module is used to construct a three-dimensional geometric skeleton of the target production equipment based on the physical structure of the target production equipment, and to bind the three-dimensional geometric skeleton to the running data base in a virtual-real synchronous manner to obtain an initial digital twin of the target production equipment. The fault anchoring module is used to anchor the operating parameter nodes in the initial digital twin with the historical fault records of the target production equipment to obtain the associated digital twin of the target production equipment. The anomaly tracing module is used to perform real-time anomaly threshold detection on the fault parameter nodes in the associated digital twin, and to perform intelligent tracing and deduction on the associated digital twin based on the detected trigger signals to obtain a preliminary diagnostic report of the target production equipment. The coupled rendering module is used to perform hierarchical coupled rendering of the preliminary diagnostic report and the associated digital twin to obtain a visual operation and maintenance interface for the target production equipment. The operation compilation module is used to compile the interactive operation behavior of the visual operation and maintenance interface into a visual operation chain to obtain precise control and maintenance instructions for the target production equipment.

Images