Historic building operation and maintenance management method and system based on digital twinning

Abstract

The application relates to the technical field of operation and maintenance management, and discloses a historical building operation and maintenance management method and system based on digital twinning, which comprises the following steps: constructing a digital basement based on historical data and point clouds, dividing a monitoring area according to a load-bearing structure, value elements and functional spaces, and associating identifiers. Running data is mapped to corresponding component storage spaces according to the belonging area, dynamic operation and maintenance archives and monitoring data streams are formed. The data streams are visually coded with the collection time as the horizontal axis and the spatial topology as the vertical axis, and a space-time distribution atlas is generated. Abnormal areas are determined according to the fluctuation form difference and trend deviation of real-time and historical data in the atlas, an abnormal list is generated by binding the identifiers and time periods. The archives are updated by taking the list as an index, the disease records are matched with the historical maintenance records, and an operation and maintenance management scheme is generated; the application can improve the efficiency of historical building operation and maintenance management based on digital twinning.

Description

Technical Field

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

[0002] Existing methods for the operation and maintenance management of historical buildings have significant shortcomings in data integration and utilization. Historical drawings, point cloud scans, repair records, and real-time sensor monitoring data are typically stored in fragmented systems or media, resulting in a fragmentation of the building's holistic information. Due to the lack of a unified digital foundation and standardized spatial indexing, it is difficult to establish precise correspondences between various operational data and specific building components. Monitoring indicators exist merely as isolated values, failing to form effective connections within the building's spatial structure and timeline. This data silo phenomenon makes it difficult for managers to intuitively grasp the real-time status and historical evolution patterns of the building as a whole and its key components when faced with massive amounts of information, reducing the data's effectiveness in supporting operational and maintenance decisions.

[0003] In terms of anomaly identification and maintenance decision-making, traditional methods rely heavily on regular manual inspections and alarm mechanisms based on single thresholds. These methods lack depth in data mining, failing to effectively distinguish between normal data fluctuations and abnormal signs, and are also ill-equipped to capture subtle trends in the long-term evolution of building structures. Once anomalies are detected, subsequent investigation, diagnosis, and maintenance plan development are often conducted in isolation, lacking a systematic framework that intelligently links abnormal areas, damage characteristics, and historical maintenance experience. This necessitates repeated investigations and analyses for each maintenance task, preventing the full utilization of past successful cases and accumulated knowledge. Consequently, operation and maintenance management efficiency is low, and the timeliness and scientific rigor of problem-solving are hampered. Summary of the Invention

[0004] This invention provides a method and system for the operation and maintenance management of historical buildings based on digital twins, the main purpose of which is to solve the problem of low efficiency in the operation and maintenance management of historical buildings based on digital twins.

[0005] To achieve the above objectives, this invention provides a method for the operation and maintenance management of historical buildings based on digital twins, comprising:

[0006] A digital base for the historical building is constructed based on historical data and point cloud data of the historical building.

[0007] Based on the load-bearing structure, distribution of value elements, and functional space of the historical building, monitoring areas are divided in the digital base, and data identifiers are associated with the monitoring areas.

[0008] Based on the monitoring areas to which the operation data of the historical buildings belong, the operation data is mapped to the storage space of the corresponding components in the digital base to obtain the dynamic operation and maintenance files of the historical buildings and the corresponding monitoring data streams.

[0009] Using the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis, the monitoring data stream is visually encoded to obtain the spatiotemporal distribution map of the historical buildings;

[0010] Based on the difference in fluctuation patterns and trend deviation between the real-time data and historical archive data in the spatiotemporal distribution map, the abnormal areas of the historical buildings are determined, and the area identifiers and abnormal time periods of the abnormal areas are bound together to obtain a list of abnormal areas of the historical buildings.

[0011] The dynamic maintenance file is updated using the area identifier in the abnormal area list as an index to obtain the current file of the historical building. The set of disease records in the current file is then matched with the historical maintenance records of the dynamic maintenance file to obtain the maintenance management plan for the historical building.

[0012] In a preferred embodiment, constructing the digital base of the historical building based on historical data and point cloud data includes:

[0013] The architectural drawings in the historical data are digitally scanned to obtain raster images of the historical buildings, and the raster images are geometrically corrected to obtain the base map of the historical buildings.

[0014] Spatially register the point cloud data with the base map to obtain the alignment point set of the historical building;

[0015] The building's outer contour line and edge feature line, which are concentrated at the alignment points, are connected according to their spatial adjacency to obtain the closed contour line of the historical building. The closed contour lines are then combined according to their spatial position within the historical building to obtain the wireframe structure of the historical building.

[0016] The repair records and material information in the historical data are used as attribute fields and attached to the corresponding component nodes in the wireframe structure to obtain the digital base of the historical building.

[0017] In a preferred embodiment, the step of dividing the monitoring area in the digital substrate based on the load-bearing structure, distribution of value elements, and functional space of the historical building, and associating the monitoring area with data identifiers, includes:

[0018] The continuous component nodes belonging to the same structural force path in the digital base are grouped into the load-bearing structural area of ​​the historical building;

[0019] The component nodes that are spatially adjacent and have the same value level in the digital base are merged into the value element area of ​​the historical building.

[0020] Based on the functional zoning boundary lines in the digital base, component nodes located within the same functional zoning are aggregated into the functional space areas of the historical building;

[0021] The boundaries of the load-bearing structure area, the value element area, and the functional space area are superimposed to obtain the comprehensive regional framework of the historical building. The overlapping part of the comprehensive regional framework is then cut and divided to obtain the defining blocks of the historical building.

[0022] A feature region code is generated for the defined block, and the feature region code is written into the metadata field contained in the spatial block within the defined block to obtain the data identifier of the spatial block.

[0023] In a preferred embodiment, the operational data of the historical building is mapped to the storage space of the corresponding component in the digital substrate based on the monitoring area to which the operational data belongs, thereby obtaining the dynamic operation and maintenance file of the historical building and the corresponding monitoring data stream, including:

[0024] Obtain the operational data package of the historical building, wherein the operational data includes the collection timestamp, spatial coordinates of the collection point, and monitoring index values;

[0025] The spatial coordinates of the acquisition points are matched with the spatial range of the monitoring area in the digital base to determine the reference monitoring area to which the running data packet belongs;

[0026] The corresponding component node space in the digital substrate is located based on the data identifier of the benchmark monitoring area;

[0027] The monitoring index values ​​are written sequentially into the monitoring data field of the component node space according to the timestamp order to obtain the independent area monitoring sequence of the historical building;

[0028] By merging the monitoring sequences of the independent areas, the monitoring data stream of the historical building is obtained;

[0029] The monitoring data stream and the real-time data fields of the corresponding component nodes are summarized to obtain the dynamic operation and maintenance file of the historical building.

[0030] In a preferred embodiment, the step of visually encoding the monitoring data stream with the acquisition time as the horizontal axis and the spatial topology of the digital substrate as the vertical axis to obtain the spatiotemporal distribution map of the historical buildings includes:

[0031] Determine the mapping rules between indicator values ​​and visual feature parameters in the monitoring data stream. The mapping rules include: determining color hue based on the value range in the monitoring data stream, determining color saturation based on the rate of change of values ​​in the monitoring data stream, and determining graphic transparency based on the cumulative value in the monitoring data stream.

[0032] Based on the area identifier of the monitoring area to which the monitoring data stream belongs, the corresponding component node in the digital base is retrieved, and the vertical axis coordinate value of the component node in the spatial topology of the digital base is read.

[0033] The index values ​​are converted into corresponding visual feature parameter values ​​according to the mapping rules;

[0034] A two-dimensional coordinate system for the historical building is constructed using the collection timestamp of the monitoring data stream as the horizontal axis and the vertical axis coordinate value as the vertical axis positioning point. Pixels are drawn point by point in the two-dimensional coordinate system using the visual feature parameter value as the pixel attribute value.

[0035] The pixels are arranged into a regular row and column structure according to the time order of the horizontal axis and the spatial order of the vertical axis to obtain the spatiotemporal two-dimensional grayscale matrix of the historical building.

[0036] The spatiotemporal two-dimensional grayscale matrix is ​​spliced ​​according to the spatial topology of the digital substrate to obtain the spatiotemporal distribution map of the historical building.

[0037] In a preferred embodiment, the step of determining the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend deviations between real-time data and historical archive data in the spatiotemporal distribution map, and binding the area identifiers and abnormal time periods of the abnormal areas to obtain a list of abnormal areas of the historical buildings, includes:

[0038] The real-time data and historical archive data of the spatiotemporal distribution map are decomposed into fluctuation patterns to obtain the number of rising segments, falling segments, peaks and troughs in the real-time data and the historical archive data, and to generate the real-time fluctuation feature vector and the historical fluctuation feature vector of the historical building.

[0039] The Euclidean distance between the real-time fluctuation feature vector and the historical fluctuation feature vector is used as the fluctuation pattern difference value of the historical building.

[0040] The difference between the linear regression slope of the real-time data and the linear regression slope of the historical archive data is used as the deviation of the trend evolution of the historical building.

[0041] The comprehensive anomaly index of the monitored area is obtained by integrating the fluctuation pattern difference value with the trend evolution deviation.

[0042] The statistical distribution characteristics of the comprehensive anomaly index are used to determine the anomaly judgment benchmark value of the historical building, and the monitoring area where the comprehensive anomaly index exceeds the anomaly judgment benchmark value is marked as the anomaly area of ​​the historical building.

[0043] In a preferred embodiment, the formula for calculating the comprehensive anomaly index includes:

[0044]

[0045] in, The comprehensive anomaly index, The weighting coefficient for the volatility pattern term. It is a natural constant. This represents the difference in fluctuation patterns. The mean of the fluctuation pattern difference. This represents the maximum difference in fluctuation patterns. It is a non-linear index with fluctuating patterns. The weighting coefficient for the trend evolution term. The deviation from the trend evolution, The standard deviation of the trend evolution. This represents the maximum deviation of the trend evolution. It is a non-linear index for trend evolution. For the weighting coefficients of the value coupling term, For comprehensive value coefficient, This represents the maximum value of the comprehensive value coefficient.

[0046] In a preferred embodiment, updating the dynamic maintenance file using the region identifier in the abnormal region list as an index to obtain the current file of the historical building, and matching the set of disease records in the current file with the historical maintenance records of the dynamic maintenance file to obtain the maintenance management plan for the historical building, includes:

[0047] Based on the region identifier in the abnormal region list, locate the corresponding monitoring region in the dynamic operation and maintenance file, and lock the component node storage space contained in the monitoring region;

[0048] Create a special exploration data field in the component node storage space, and write the start and end identifiers of the abnormal time periods in the abnormal area list into the metadata area of ​​the special exploration data field;

[0049] Obtain a special survey data package for the component nodes during the abnormal period, the special survey data package including the physicochemical analysis results of material samples and the path of high-definition image files;

[0050] The feature data in the special survey data package is written into the corresponding subfields of the special survey data field to complete the data update of the component node and obtain the current archive of the historical building.

[0051] The disease type labels and disease severity levels in the existing disease record set are matched with the historical maintenance records in the dynamic operation and maintenance archive, and the matched historical maintenance measures are associated with the corresponding component node information to obtain the operation and maintenance management plan for the historical building.

[0052] In a preferred embodiment, the step of matching the disease type tags and disease severity levels in the current disease record set with the historical maintenance records in the dynamic operation and maintenance archive, and then associating and summarizing the matched historical maintenance measures with the corresponding component node information to obtain the operation and maintenance management plan for the historical building, includes:

[0053] Retrieve historical maintenance records with the same disease type tag from the historical maintenance record database of the dynamic operation and maintenance archive to obtain a preliminary matching record set of the historical building;

[0054] The optimal matching record for the historical building is obtained by selecting the historical maintenance record with the smallest difference from the severity level of the disease from the preliminary matching record set.

[0055] The maintenance measures text and maintenance effect evaluation data in the optimal matching record are associated with the component node information corresponding to the disease record set to obtain the maintenance management plan for the historical building.

[0056] To address the above problems, the present invention also provides a historical building operation and maintenance management system based on digital twins, the system comprising:

[0057] The digital infrastructure module constructs a digital infrastructure for the historical building based on historical data and point cloud data of the historical building.

[0058] The monitoring area module divides the monitoring area in the digital base according to the load-bearing structure, value element distribution and functional space of the historical building, and associates the monitoring area with data identifiers.

[0059] The operation and maintenance file module maps the operation data of the historical building to the storage space of the corresponding component in the digital base based on the monitoring area to which the operation data belongs, so as to obtain the dynamic operation and maintenance file of the historical building and the corresponding monitoring data stream.

[0060] The spatiotemporal distribution mapping module uses the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis to visually encode the monitoring data stream, thereby obtaining the spatiotemporal distribution map of the historical building.

[0061] The abnormal area module determines the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend evolution deviations between the real-time data and historical archive data in the spatiotemporal distribution map, and binds the area identifier and abnormal time period of the abnormal areas to obtain a list of abnormal areas of the historical buildings.

[0062] The operation and maintenance management solution module updates the dynamic operation and maintenance file using the area identifier in the abnormal area list as an index to obtain the current file of the historical building, and matches the set of disease records in the current file with the historical maintenance records of the dynamic operation and maintenance file to obtain the operation and maintenance management solution for the historical building.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1. This digital twin-based historical building operation and maintenance management method constructs a digital foundation integrating geometric structure and historical information, and divides monitoring areas into multiple dimensions based on load-bearing structure, value elements, and functional spaces, achieving refined organization and spatial indexing of historical building data. This method precisely maps real-time operational data to the storage space of corresponding component nodes according to spatial location, forming dynamically updated operation and maintenance archives. Furthermore, it generates spatiotemporal distribution maps through visual coding technology, allowing for a direct presentation of the evolution patterns of monitoring indicators in both time and space dimensions. This significantly improves the management efficiency and readability of massive amounts of heterogeneous data, providing a highly structured and visualized data foundation for subsequent analysis.

[0065] 2. This method, through quantitative analysis of the differences in fluctuation patterns and trend deviations between real-time data and historical archive data in the spatiotemporal distribution map, can accurately identify abnormal areas and their occurrence periods, and bind abnormal information with area identifiers to form an abnormal area list. Based on this, the archives are dynamically updated and matched with historical maintenance records using abnormal areas as indexes, enabling the rapid generation of operation and maintenance management plans tailored to the current type and severity of damage. This achieves a data-driven process from anomaly early warning and on-site investigation to plan formulation, significantly improving the accuracy of anomaly response and the scientific nature of maintenance decisions, ensuring the effectiveness and relevance of historical building protection measures. Attached Figure Description

[0066] Figure 1 A flowchart illustrating a method for the operation and maintenance management of historical buildings based on digital twins, provided in an embodiment of the present invention;

[0067] Figure 2 A functional module diagram of a digital twin-based historical building operation and maintenance management system provided in an embodiment of the present invention;

[0068] 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

[0069] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0070] This application provides a method for the operation and maintenance management of historical buildings 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, the method for the operation and maintenance management of historical buildings based on digital twins 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 cloud server cluster. 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.

[0071] Reference Figure 1 The diagram shown is a flowchart illustrating a historical building operation and maintenance management method based on digital twins according to an embodiment of the present invention. In this embodiment, the historical building operation and maintenance management method based on digital twins includes:

[0072] In this embodiment of the invention, the construction of the digital base of the historical building based on historical data and point cloud data is specifically used for:

[0073] The architectural drawings in the historical data are digitally scanned to obtain raster images of the historical buildings, and the raster images are geometrically corrected to obtain the base map of the historical buildings.

[0074] Spatially register the point cloud data with the base map to obtain the alignment point set of the historical building;

[0075] The building's outer contour line and edge feature line, which are concentrated at the alignment points, are connected according to their spatial adjacency to obtain the closed contour line of the historical building. The closed contour lines are then combined according to their spatial position within the historical building to obtain the wireframe structure of the historical building.

[0076] The repair records and material information in the historical data are used as attribute fields and attached to the corresponding component nodes in the wireframe structure to obtain the digital base of the historical building.

[0077] Specifically, the paper architectural drawings of historical buildings are scanned, with an appropriate scanning resolution set to ensure that the lines and text on the drawings are clearly legible. After scanning, the drawings are saved as lossless compressed raster images. At least four feature points evenly distributed throughout the drawings are selected on the raster image, such as the corners of walls or the corners of door and window openings. The real-world three-dimensional coordinates of these feature points are measured on-site, or their coordinate values ​​are obtained from existing precise surveying data.

[0078] Specifically, the point cloud data contains densely packed three-dimensional coordinate points on the surface of historical buildings. Corresponding points on the base map are manually or automatically identified within the point cloud, such as significant corners of the building facade and eaves endpoints. An iterative nearest-point algorithm is employed to continuously calculate the nearest-point correspondence between the point cloud and the base map, thereby determining the optimal rotation matrix and translation vector, ensuring that the corresponding points in the point cloud precisely coincide with their corresponding points on the base map in spatial location.

[0079] Specifically, a point set belonging to the building's outer contour and edge features, such as wall boundary points, roof ridge points, and door and window opening edge points, is separated from the aligned point cloud using a point cloud segmentation algorithm. These contour points are then sorted, and adjacent points are connected sequentially according to the principle of closest spatial Euclidean distance to form a series of continuous polylines. For contours that constitute closed regions, such as wall cross-section contours, the beginning and end points of the polylines are connected to generate closed contour lines.

[0080] Specifically, the repair records in the historical archives are compiled, including detailed textual information such as the time, reason, construction technique, and construction unit of each repair, as well as descriptive information such as the type, source, and properties of materials used in each part of the building. For each component node in the wireframe structure, such as each wall, column, and floor slab, an independent attribute storage area is allocated in the data structure.

[0081] Furthermore, using these feature points as control points, a mapping relationship between image pixel coordinates and real-world coordinates is established through a polynomial transformation model. The entire raster image is resampled to eliminate geometric distortions caused by paper deformation and scanning angle deviations, ensuring that every detail in the image accurately corresponds to its actual spatial location. Finally, a geometrically corrected baseline map is obtained, which serves as a unified reference for aligning all subsequent spatial data.

[0082] Furthermore, after multiple iterations of optimization until the registration error is less than the preset allowable range, the transformed point cloud coordinate system is then converted to the coordinate system of the base map. At this point, the point cloud data and the base map achieve spatial consistency, resulting in an aligned point cloud set.

[0083] Furthermore, based on the actual positions of these closed contour lines in three-dimensional space, such as according to different floor elevations and different facade partitions, the contour lines belonging to the same structural component are combined. For example, the inner and outer closed contour lines of the same wall are combined to form the wireframe of the wall, and the bottom and side contour lines of the same beam are combined to form the wireframe of the beam. Through such combination operations, the wireframe structural model of the entire building is gradually constructed. This model accurately expresses the geometric shape and spatial relationship of each component of the building.

[0084] Furthermore, based on the component's location and number within the building, corresponding repair records and material information are added to the component's attribute fields in key-value pairs. For example, a repair history field is added to a wall segment, containing the text content of each repair record, and a main material field is added, with the value being the specific material name. After all component attributes are attached, the wireframe structure is upgraded from a simple geometric model to a digital model incorporating rich historical information—the final baseline drawing.

[0085] In summary, converting historical paper drawings into digital images with precise spatial references eliminates geometric distortions introduced by the original drawings due to their age, paper deformation, or scanning process. This provides a unified and accurate coordinate benchmark for the alignment and fusion of all subsequent spatial data, ensuring the reliability of the digital base in terms of spatial accuracy.

[0086] In summary, the precise spatial alignment of point cloud data obtained from high-precision 3D scanning with geometrically calibrated baseline maps ensures that the detailed 3D geometric information carried by the point cloud strictly corresponds to the 2D contours and historical information in the drawings. This lays the data foundation for accurately extracting the structural contours and feature lines of buildings from the point cloud and achieves the effective fusion of multi-source heterogeneous data.

[0087] In summary, the discrete point cloud data is transformed into a wireframe structural model with clear topological relationships, consisting of continuous closed contour lines. This model not only accurately expresses the geometric shape and spatial relationship of each building component, but also provides a structured carrier for subsequent attribute information, enabling building information to be organized and managed by component.

[0088] In summary, non-geometric information originally scattered in historical archives was linked one by one with the specific component nodes in the wireframe structure, so that the digital model not only has a geometric form, but also carries rich historical and attribute information, thus completing the sublimation from a pure geometric model to an information model. The resulting baseline map provides a data foundation with a complete historical background for subsequent monitoring, analysis and management based on digital twins.

[0089] In this embodiment of the invention, the step of dividing the monitoring area in the digital substrate based on the load-bearing structure, value element distribution, and functional space of the historical building, and associating the monitoring area with data identifiers, is specifically used for:

[0090] The continuous component nodes belonging to the same structural force path in the digital base are grouped into the load-bearing structural area of ​​the historical building;

[0091] The component nodes that are spatially adjacent and have the same value level in the digital base are merged into the value element area of ​​the historical building.

[0092] Based on the functional zoning boundary lines in the digital base, component nodes located within the same functional zoning are aggregated into the functional space areas of the historical building;

[0093] The boundaries of the load-bearing structure area, the value element area, and the functional space area are superimposed to obtain the comprehensive regional framework of the historical building. The overlapping part of the comprehensive regional framework is then cut and divided to obtain the defining blocks of the historical building.

[0094] A feature region code is generated for the defined block, and the feature region code is written into the metadata field contained in the spatial block within the defined block to obtain the data identifier of the spatial block.

[0095] Specifically, in the digital foundation, all component nodes are first traversed to identify the structural function of each node, such as beams, columns, and load-bearing walls. Based on the principles of structural mechanics, the main paths for load transfer from the roof and floors to the foundation are determined, and these paths consist of a series of interconnected component nodes.

[0096] Specifically, in the digital base, all component nodes are traversed, and the pre-stored value assessment attribute value of each node is read. This attribute value is determined according to the value assessment standards of historical buildings. For example, components with extremely high historical value are marked as first-level value, and components with relatively high historical value are marked as second-level value.

[0097] Specifically, in the digital infrastructure, based on the functional zoning design drawings of the historical building or the results of the current usage survey, clear functional zoning boundary lines are drawn in advance in the spatial coordinate system of the digital infrastructure. These boundary lines are closed polylines that divide the interior space of the building into different functional areas, such as exhibition areas, office areas, storage areas, and traffic areas.

[0098] Specifically, during the overlay process, the original boundaries of all areas are preserved, and these boundaries are merged and calculated in three-dimensional space to form a comprehensive regional framework map covering the entire historical building, composed of multiple complex boundary lines. In this comprehensive regional framework map, the originally independently divided regional boundaries intersect with each other, creating numerous overlapping spatial blocks.

[0099] Specifically, the coding rules adopt a hierarchical combination approach. First, the building floor number or zone number where the defined block is located is determined based on its position in the overall building space as the first part of the code. Then, the structural type code is determined based on the main load-bearing structural characteristics of the block as the second part of the code. Next, the value level code is determined based on the level of the main value elements it contains as the third part of the code. Finally, the functional type code is determined based on the functional space type to which it belongs as the fourth part of the code. These four parts are combined with connectors to form a complete feature area code.

[0100] Furthermore, by analyzing the connections between nodes, adjacent nodes located on the same load transfer path are connected in series to form a continuous set of nodes. For example, starting from the main beam node of the roof, it connects downwards to the column node supporting it, then to the beam node below, and finally to the foundation node. These nodes together constitute a complete structural force path. All continuous sets of nodes belonging to different force paths are marked separately, and each such set is an independent load-bearing structural region.

[0101] Furthermore, a spatial neighborhood search method is employed to find directly adjacent nodes in space around each node with the same value level. These adjacent nodes may be connected by shared faces, shared edges, or very short distances. All spatially adjacent nodes with the same value level are gradually merged to form a continuous area of ​​value elements. For example, an area formed by merging multiple adjacent exquisite wood carvings of the same first-level value, or an area formed by merging multiple consecutive doors and windows with the same historical value.

[0102] Furthermore, all component nodes in the digital substrate are traversed, and each node's spatial coordinates are determined to be within a functional zone boundary line. A spatial determination method based on point-polygon relationships is used: a ray is emitted from the component node's coordinates in any direction, and the number of intersections between this ray and the functional zone boundary line is calculated. If the number of intersections is odd, the node is located inside the boundary line and is determined to belong to that functional zone. All component nodes located within the same functional zone boundary line are grouped together to form a functional space area. For example, all component nodes such as walls, floors, and display cases located within the exhibition area boundary line collectively constitute the exhibition functional space area.

[0103] Furthermore, these overlapping blocks are divided by breaking the boundary lines at all intersections. Then, the regions are regenerated based on the broken line segments, so that each newly generated region has a unique boundary combination and no longer overlaps with other regions. These new regions, which are obtained after cutting and have independent and non-overlapping spatial ranges, are the defining blocks of historical buildings. Each defining block corresponds to a uniquely defined part in space.

[0104] Furthermore, in the data structure of the digital substrate, each defined block corresponds to a spatial block. The metadata storage area of ​​the spatial block is located, and the generated feature region code is written into it as a piece of metadata. This feature region code becomes the data identifier of the spatial block, which is used to uniquely identify and index the spatial block in subsequent operation and maintenance management.

[0105] In summary, by extracting the key structural system that bears the overall load of a building from a complex architectural model, and reorganizing the dispersed beams, columns, walls and other components according to the mechanical transmission path, load-bearing areas with clear physical meaning are formed. This allows subsequent monitoring and maintenance to prioritize the parts that have a decisive impact on building safety, and provides targeted spatial units for structural safety assessment.

[0106] In general, spatially clustering components with the same level of protection value scattered throughout the building creates continuous areas of value elements. This prevents high-value components from being viewed in isolation, allowing them to be managed and protected as a whole. It also facilitates the development of differentiated monitoring frequencies and protection strategies for areas with different value levels, ensuring the overall preservation of historical value.

[0107] In summary, by combining architectural models with actual functional uses and regrouping components according to different functional requirements such as exhibition, office, and storage, subsequent operational data can correspond to specific usage scenarios. This provides an organizational framework for assessing the environmental adaptability and usage intensity of different functional spaces and for developing function-based operation and maintenance measures.

[0108] In summary, the three division systems based on mechanics, value, and function are spatially integrated. By overlapping and dividing boundaries, the boundary conflicts between different division standards are eliminated, generating the smallest management unit that is unique and non-overlapping in space, namely the defined block. Each block simultaneously possesses structural, value, and functional attributes, laying a spatial foundation for subsequent multi-dimensional and refined data management.

[0109] In summary, each spatially unique delineation block is assigned a globally unique identifier. This code comprehensively reflects the block's structure, value, and functional characteristics, achieving a precise correspondence between spatial units and data records. This enables all subsequent monitoring data, maintenance records, and management information to be quickly indexed and associated through this identifier, providing a crucial data retrieval method for building traceable and queryable dynamic operation and maintenance archives.

[0110] In this embodiment of the invention, when mapping the operational data of the historical building to the storage space of the corresponding component in the digital substrate based on the monitoring area to which the operational data belongs, and obtaining the dynamic operation and maintenance file of the historical building and the corresponding monitoring data stream, it is specifically used for:

[0111] Obtain the operational data package of the historical building, wherein the operational data includes the collection timestamp, spatial coordinates of the collection point, and monitoring index values;

[0112] The spatial coordinates of the acquisition points are matched with the spatial range of the monitoring area in the digital base to determine the reference monitoring area to which the running data packet belongs;

[0113] The corresponding component node space in the digital substrate is located based on the data identifier of the benchmark monitoring area;

[0114] The monitoring index values ​​are written sequentially into the monitoring data field of the component node space according to the timestamp order to obtain the independent area monitoring sequence of the historical building;

[0115] By merging the monitoring sequences of the independent areas, the monitoring data stream of the historical building is obtained;

[0116] The monitoring data stream and the real-time data fields of the corresponding component nodes are summarized to obtain the dynamic operation and maintenance file of the historical building.

[0117] Specifically, raw monitoring data is received in real time or periodically from sensor networks, IoT data collectors, or manual inspection terminals deployed in various monitoring areas of the historical building. This data is then encapsulated into standard format data packets according to a predefined data structure.

[0118] Specifically, each monitoring region in the digital substrate is stored in the data structure as the spatial extent enclosed by its boundary polygon. For each acquisition point coordinate, the inclusion test between the point and the polygon is performed using the ray casting method.

[0119] Specifically, based on the data identifiers of the baseline monitoring areas obtained by matching the running data packets, precise retrieval is performed within the data structure of the digital substrate. The digital substrate employs a tree-like or graph-like data organization method, with the data identifier of each monitoring area serving as the primary key, establishing a mapping relationship with the corresponding component node storage space.

[0120] Specifically, all operational data packets belonging to the same baseline monitoring area are sorted according to the order of their collection timestamps. After sorting, these ordered data packets are traversed sequentially to extract the monitoring indicator values. For each monitoring indicator value, a new data record is created in the monitoring data field of the component node space based on its collection timestamp. This record contains both the timestamp and the indicator value.

[0121] Specifically, independent monitoring sequences are obtained from each monitoring area in the digital substrate, and these sequences are distributed across different component node spaces. To form a global monitoring data view, all independent monitoring sequences are globally merged and sorted according to their acquisition timestamps, constructing a unified time series covering all monitoring points throughout the entire historical building.

[0122] Specifically, the historical building monitoring data stream generated in the previous step is integrated with the real-time data fields of all component nodes in the digital infrastructure. The real-time data field is a reserved data area within each component node space for storing the latest monitoring status, such as the latest monitoring value, data update time, and warning status. By traversing each component node in the digital infrastructure, the content of its real-time data field is updated based on the latest record of the corresponding node in the monitoring data stream.

[0123] Furthermore, each running data package contains three core elements: a data acquisition timestamp, which precisely records the moment the data was generated or acquired, in a standardized time string including year, month, day, hour, minute, and second; spatial coordinates of the acquisition point, which represent the location of the monitoring point in actual space in three-dimensional coordinate values, based on the same spatial reference system as the digital substrate; and monitoring index values, which are the specific values ​​of physical quantities measured by the sensor, such as vibration frequency, tilt angle, crack width, temperature, and humidity values.

[0124] Furthermore, draw a ray from this point in any direction and calculate the number of intersections between the ray and the polygonal boundary of the monitoring area. If the number of intersections is odd, the point is located inside the polygon, and the coordinates of the acquisition point are determined to fall within this monitoring area. Traverse all monitoring areas until a unique monitoring area containing the coordinates of the acquisition point is found. This area is then identified as the baseline monitoring area to which this running data packet belongs, and the unique identifier of this baseline monitoring area is recorded and associated with this data packet.

[0125] Furthermore, by constructing a hash index and directly searching using the data identifier as the key, the storage space address within the digital matrix that stores all component data within the monitoring area can be instantly located. This component node space is a memory region or database record set specifically used to store the geometry, attributes, and monitoring data of each component node belonging to this monitoring area. Once the location is successful, a channel for writing data to this component node space is opened.

[0126] Furthermore, the monitoring data fields adopt a time-series database storage method, with new records being appended to the monitoring sequence corresponding to each component node. As data is continuously written, each component node in the monitoring area forms a time-ordered sequence of monitoring values, which together constitute an independent regional monitoring sequence for that monitoring area.

[0127] Furthermore, during the merging process, the original regional identifiers and component node information of each monitoring data point are retained to ensure data traceability. The long sequence obtained after merging and sorting is the monitoring data stream that reflects the changes of various parts of the historical building over time as a whole. This data stream provides a unified data source for subsequent analysis and visualization.

[0128] Furthermore, the static attribute data of all component nodes, historical maintenance records, real-time monitoring data, and the complete monitoring data stream are organized into a unified archive database to form a dynamic operation and maintenance archive for historical buildings. This archive uses component nodes as the basic unit and fully records comprehensive data from the building's historical information to its current real-time status, supporting queries and analyses by time, region, component, and other dimensions.

[0129] In summary, it standardizes the unified format of raw data collected by various sensors and monitoring equipment, and encapsulates the three core elements of time, space and numerical values ​​in a structured manner. This provides a standardized data carrier for monitoring data from different sources and of different types, and provides a complete and resolvable data source for all subsequent time- and space-based data processing and analysis, eliminating integration barriers caused by differences in data formats.

[0130] In summary, it achieves a precise correspondence between real-time monitoring data in the physical world and virtual spatial locations in the digital twin model. By judging spatial inclusion relationships, discrete collection point data is assigned to pre-divided monitoring areas, ensuring that each data packet can be correctly classified to its actual spatial location, providing a spatial index basis for subsequent data storage, retrieval, and analysis by region.

[0131] In summary, the system establishes a clear path from the monitoring area to the specific component storage unit. By using data identifiers as retrieval keys, it can quickly and accurately locate the physical memory or database location in the digital substrate responsible for storing the component data of that area. This enables precise transitions from macro-level area division to micro-level component storage, establishing an efficient access channel for subsequent refined data writing.

[0132] In summary, by organizing continuous monitoring data according to component nodes, a strictly chronological sequence of monitoring values ​​is formed in the storage space of each component node. This sequence fully records the state evolution of the component on the time axis, providing a structured data foundation for subsequent analysis of the long-term change patterns and fluctuation characteristics of individual components, as well as for time series analysis.

[0133] In summary, by globally merging the independent time series data scattered across various component nodes, a unified time series data stream covering all monitoring points of the entire historical building is formed. This data stream breaks down the spatial isolation between components and presents the global trend of monitoring indicators evolving over time from the perspective of the building as a whole, providing a complete data view for subsequent spatiotemporal analysis and anomaly detection across the entire area.

[0134] In summary, a deep integration of static historical information and dynamic real-time monitoring data has been achieved. The global monitoring data stream and the latest status data of each component node have been integrated into a unified archive structure, forming a comprehensive information database that has both historical background and real-time status. This dynamic operation and maintenance archive can continuously update itself with the arrival of new data, providing comprehensive, accurate and timely data support for subsequent anomaly diagnosis, maintenance decisions and value assessment.

[0135] In this embodiment of the invention, when visually encoding the monitoring data stream with the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis to obtain the spatiotemporal distribution map of the historical buildings, the specific method is as follows:

[0136] Determine the mapping rules between indicator values ​​and visual feature parameters in the monitoring data stream. The mapping rules include: determining color hue based on the value range in the monitoring data stream, determining color saturation based on the rate of change of values ​​in the monitoring data stream, and determining graphic transparency based on the cumulative value in the monitoring data stream.

[0137] Based on the area identifier of the monitoring area to which the monitoring data stream belongs, the corresponding component node in the digital base is retrieved, and the vertical axis coordinate value of the component node in the spatial topology of the digital base is read.

[0138] The index values ​​are converted into corresponding visual feature parameter values ​​according to the mapping rules;

[0139] A two-dimensional coordinate system for the historical building is constructed using the collection timestamp of the monitoring data stream as the horizontal axis and the vertical axis coordinate value as the vertical axis positioning point. Pixels are drawn point by point in the two-dimensional coordinate system using the visual feature parameter value as the pixel attribute value.

[0140] The pixels are arranged into a regular row and column structure according to the time order of the horizontal axis and the spatial order of the vertical axis to obtain the spatiotemporal two-dimensional grayscale matrix of the historical building.

[0141] The spatiotemporal two-dimensional grayscale matrix is ​​spliced ​​according to the spatial topology of the digital substrate to obtain the spatiotemporal distribution map of the historical building.

[0142] Specifically, the first step is to statistically analyze all indicator values ​​in the monitoring data stream to determine the entire range of variation for each indicator. This range is then divided into multiple consecutive numerical intervals. For example, the vibration amplitude can be divided into low-frequency, mid-frequency, and high-frequency intervals from minimum to maximum, or even more finely. Each numerical interval is assigned a specific color hue; for example, blue corresponds to the low-frequency interval, green to the mid-frequency interval, and red to the high-frequency interval. In this way, the magnitude of the monitoring indicator value can be directly represented by the color hue.

[0143] Specifically, to determine the vertical axis position of each pixel in the spatiotemporal map, it is necessary to obtain the spatial height information corresponding to each monitoring data stream. Based on the region identifier of the monitoring area to which the monitoring data stream belongs, all component nodes contained in that monitoring area are directly located in the data structure of the digital substrate using a hash retrieval method.

[0144] Specifically, for each monitoring record in the monitoring data stream, that is, the index value corresponding to each collection moment, the mapping rules determined in the first step are applied to convert the visual feature parameters. First, it is determined which pre-divided value interval the index value falls into, and the color hue of the record is determined according to the correspondence between the interval and hue.

[0145] Specifically, a two-dimensional coordinate system is constructed for drawing spatiotemporal maps, with the acquisition timestamp as the horizontal axis and the component node coordinates read from the digital substrate as the vertical axis. In this two-dimensional plane, for each monitoring record in the monitoring data stream, the horizontal coordinate position is determined based on its acquisition timestamp, and the vertical coordinate position is determined based on the vertical coordinate value of its corresponding component node, thereby determining a unique pixel drawing point.

[0146] Specifically, along the horizontal axis, all pixels are arranged from left to right at equal intervals according to the order of their acquisition timestamps to ensure temporal continuity. Along the vertical axis, all pixels are arranged from low to high according to the vertical coordinate values ​​read from the digital substrate to ensure spatial order.

[0147] Specifically, the spatial topology of the digital foundation includes not only height information but also the horizontal layout of the building, such as the spatial relationships between different rooms and facades. Based on these horizontal spatial relationships, spatiotemporal two-dimensional grayscale matrices corresponding to different horizontal positions are stitched together. For example, the grayscale matrices of the east, south, west, and north facades are combined according to their actual orientation, or the grayscale matrices of the first, second, and third floors are stacked vertically according to the floor order.

[0148] Furthermore, for each value in the monitoring data stream at each time point, its rate of change relative to the previous time point is calculated. The rate of change is determined by comparing the difference between values ​​at two adjacent times, and this rate of change is mapped to the intensity of color saturation. A higher rate of change results in higher color saturation and a more vibrant color, while a lower rate of change results in lower saturation and a duller color. Finally, for each monitoring point in the monitoring data stream, the cumulative value from the start of monitoring to the current time is calculated, such as cumulative deformation or cumulative vibration energy. The magnitude of this cumulative value is mapped to the intensity of graphic transparency. A larger cumulative value results in lower transparency and a less transparent graphic, while a smaller cumulative value results in higher transparency and a more transparent graphic. These three mapping rules together determine the complete conversion basis from indicator values ​​to visual feature parameters.

[0149] Furthermore, the spatial topology of the digital substrate predefines a vertical coordinate value for each component node. This vertical coordinate value can be the height of the component node's center point or the height of the floor where the component node is located relative to the building's reference plane, representing the component's position in the building's vertical direction. This vertical coordinate value is read from the located component node and used as the vertical positioning point for subsequently plotting the monitoring data stream pixels in a two-dimensional coordinate system, ensuring that monitoring data at different heights are arranged in the map according to their actual spatial height.

[0150] Further, the difference between the index value and the value at the adjacent previous time step is calculated. The color saturation is determined based on the magnitude of the difference; the larger the difference, the higher the saturation. Next, all index values ​​for this monitoring point from the start of monitoring to the current time are accumulated to obtain the cumulative value at the current time. The image transparency is determined based on the magnitude of the cumulative value; the larger the cumulative value, the lower the transparency. After these three steps, a complete set of visual feature parameter values ​​is generated for the index value at each monitoring time step, including the determined hue, saturation, and transparency. These parameter values ​​will be used for subsequent pixel rendering.

[0151] Furthermore, the visual feature parameter values ​​obtained from the mapping transformation of this monitoring record, namely hue, saturation, and transparency, are combined to form the color attribute value of that pixel. In a two-dimensional coordinate system, each pixel is drawn point by point according to the determined horizontal and vertical coordinate positions, and each pixel represents a visual representation of the monitoring indicator value at a specific time and a specific height position.

[0152] Furthermore, after this arrangement, the originally discrete pixels are organized into a two-dimensional matrix with a fixed number of rows and columns. Each element in the matrix corresponds to the color value of a pixel. This matrix is ​​the spatiotemporal two-dimensional grayscale matrix, which completely records the distribution of the monitoring indicators over the entire time period and the entire spatial height.

[0153] Furthermore, through this splicing operation, a spatiotemporal distribution map that comprehensively reflects the overall spatiotemporal changes of historical buildings is finally generated. This map intuitively displays the evolution of monitoring indicators in both time and space dimensions in the form of a two-dimensional image.

[0154] In summary, a quantitative conversion relationship between the physical values ​​of monitoring indicators and the visual elements of images was established. The hue of the color is determined by the range of values, allowing observers to quickly determine the range of values ​​by the type of color. The saturation of the color is determined by the rate of change of the values, allowing the intensity of change to be intuitively reflected by the vividness of the colors. The transparency of the image is determined by the cumulative amount of values, allowing the long-term cumulative effect to be perceived by the transparency of the image. This mapping rule provides a complete coding basis for the subsequent transformation of abstract multidimensional monitoring data into intuitive two-dimensional images.

[0155] In summary, it achieves precise correlation between monitoring data streams and their spatial height locations, quickly locates specific component nodes using area identifiers, and extracts predefined vertical axis coordinate values ​​from them. These vertical axis coordinate values ​​represent the actual height of the component in the vertical direction of the building, providing a spatial positioning benchmark for the subsequent accurate arrangement of monitoring data at different height positions in a two-dimensional coordinate system. This ensures that the spatiotemporal map can truly reflect the distribution characteristics of monitoring indicators as height changes.

[0156] In summary, the index values ​​at each time point in the monitoring data stream are converted into visual feature parameter values ​​composed of hue, saturation, and transparency according to a predetermined mapping rule. This conversion process enables the physical monitoring data, which was originally impossible to observe directly, to acquire color attributes that can be processed and rendered by a computer graphics system. This provides directly usable pixel color values ​​for drawing pixels in a two-dimensional coordinate system, thus completing the digital mapping from the physical world to the visual world.

[0157] In summary, by integrating information from both time and space dimensions with the visual characteristics of monitoring indicators, the horizontal coordinate of each pixel in the constructed two-dimensional plane corresponds to the time of data collection, the vertical coordinate corresponds to the height position of the component, and the color of the pixel carries the numerical characteristics of the monitoring indicator at that time and position. This drawing process condenses multidimensional heterogeneous data into a single image element, realizing the visual mapping of monitoring information in a two-dimensional spatiotemporal space.

[0158] In summary, the discretely drawn pixels are arranged in a strict temporal and spatial order to form a regular matrix structure with a fixed number of rows and columns. The rows of the matrix correspond to different spatial height positions, the columns correspond to different acquisition time points, and each element value in the matrix is ​​the color value of the corresponding pixel. This regularized data organization provides standardized data input for subsequent computer image processing, pattern recognition, and quantitative analysis.

[0159] In summary, multiple spatiotemporal two-dimensional grayscale matrices representing different building facades or floors are combined and stitched together based on the real spatial relationships recorded in the digital substrate. This results in a comprehensive image that fully reflects the changes in monitoring indicators of the entire historical building in both time and space. This spatiotemporal distribution map allows managers to clearly observe the spatial distribution of abnormal areas, the time range of abnormal occurrences, and the evolution trend of abnormalities over time and space. It provides an intuitive and efficient decision support tool for quickly locating problems, analyzing causes, and formulating maintenance strategies.

[0160] In this embodiment of the invention, the step of determining the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend deviations between real-time data and historical archive data in the spatiotemporal distribution map, and binding the area identifiers and abnormal time periods of the abnormal areas to obtain a list of abnormal areas of the historical buildings, is specifically used for:

[0161] The real-time data and historical archive data of the spatiotemporal distribution map are decomposed into fluctuation patterns to obtain the number of rising segments, falling segments, peaks and troughs in the real-time data and the historical archive data, and to generate the real-time fluctuation feature vector and the historical fluctuation feature vector of the historical building.

[0162] The Euclidean distance between the real-time fluctuation feature vector and the historical fluctuation feature vector is used as the fluctuation pattern difference value of the historical building.

[0163] The difference between the linear regression slope of the real-time data and the linear regression slope of the historical archive data is used as the deviation of the trend evolution of the historical building.

[0164] The comprehensive anomaly index of the monitored area is obtained by integrating the fluctuation pattern difference value with the trend evolution deviation.

[0165] The statistical distribution characteristics of the comprehensive anomaly index are used to determine the anomaly judgment benchmark value of the historical building, and the monitoring area where the comprehensive anomaly index exceeds the anomaly judgment benchmark value is marked as the anomaly area of ​​the historical building.

[0166] Specifically, real-time time series data and historical archive time series data for the corresponding historical periods are extracted from the spatiotemporal distribution map for each monitoring area, and fluctuation pattern decomposition is performed on these two series respectively. During decomposition, every data point in the entire time series is traversed first, and the value of the current point is compared with the values ​​of the adjacent points before and after it. Peaks and the number of peaks are identified by detecting local extreme points: if the value of a point is greater than the values ​​of its two adjacent points before and after it, the point is marked as a peak; if the value of a point is less than the values ​​of its two adjacent points before and after it, the point is marked as a trough.

[0167] Specifically, the difference between the real-time fluctuation feature vector obtained in the previous step and the historical fluctuation feature vector is measured, using Euclidean distance as the difference value of fluctuation pattern. In practice, the two vectors are regarded as two points in four-dimensional space. First, the difference between the two vectors in each dimension is calculated, namely, the number of real-time rising segments minus the number of historical rising segments, the number of real-time falling segments minus the number of historical falling segments, the number of real-time peaks minus the number of historical peaks, and the number of real-time troughs minus the number of historical troughs, resulting in four differences.

[0168] Specifically, linear regression analysis was performed on real-time time-series data and historical archive time-series data for the same monitoring area to obtain their respective trend slopes. The purpose of linear regression is to find a straight line that best represents the overall trend of data change, i.e., the linear relationship between time and indicator values. First, the average value of all time points and the average value of all indicator values ​​in the real-time data and historical archive data were calculated respectively.

[0169] Specifically, the calculated fluctuation pattern difference value and trend evolution deviation are fused to obtain a comprehensive anomaly index for each monitoring area. The fusion adopts a weighted summation method. First, a weight coefficient is assigned to both the fluctuation pattern difference value and the trend evolution deviation. These two weight coefficients are pre-set based on factors such as the structural characteristics of historical buildings and the importance of monitoring indicators.

[0170] Specifically, a dataset is constructed by collecting the comprehensive anomaly index of all monitored areas at the current moment. The statistical distribution characteristics of this dataset are then calculated. For example, the mean and standard deviation of all comprehensive anomaly indices are calculated, and the mean plus twice the standard deviation is used as the anomaly judgment benchmark. Alternatively, the percentile method is used to sort the data from smallest to largest and take the corresponding percentile as the anomaly judgment benchmark.

[0171] Furthermore, while marking peaks and troughs, the total number of marked peaks and troughs in the entire sequence is counted. Next, the time series is divided into continuous rising and falling segments based on the monotonic changes in values: starting from the beginning of the sequence, the values ​​of two adjacent points are compared. If the later point is greater than the earlier point, it enters a rising segment, continuing until a falling segment appears. Each continuous rising region is counted as a rising segment; similarly, a continuous falling region is counted as a falling segment. After traversing the entire sequence, the total number of rising segments and the total number of falling segments are obtained. These four statistics—the number of rising segments, the number of falling segments, the number of peaks, and the number of troughs—are combined sequentially into a four-dimensional vector, which is the fluctuation feature vector of this time series. Real-time data corresponds to a real-time fluctuation feature vector, and historical archive data corresponds to a historical fluctuation feature vector.

[0172] Furthermore, each difference is squared to obtain four squared values, which are then added together to obtain a sum. Finally, the square root of this sum is taken, and the resulting value is the fluctuation pattern difference value. This value quantitatively reflects the overall deviation between the real-time fluctuation pattern of the current monitoring area and the typical historical fluctuation pattern.

[0173] Furthermore, for each point in time in the real-time data, the difference between that point and the time average, and the difference between the indicator value at that moment and the indicator average, are calculated. These two differences are multiplied and summed to obtain the numerator. Simultaneously, the difference between each point in time and the time average is squared and summed to obtain the denominator. The numerator divided by the denominator yields the linear regression slope of the real-time data. The linear regression slope of the historical archive data is calculated using the exact same method. Finally, the linear regression slope of the real-time data is subtracted from the linear regression slope of the historical archive data; the resulting difference is the trend deviation, which reflects the direction and magnitude of the current monitoring area's trend deviation relative to the historical long-term trend.

[0174] Furthermore, the fluctuation pattern difference value of each monitoring area is multiplied by its weight to obtain the weighted fluctuation difference value; then, the trend evolution deviation of each monitoring area is multiplied by its weight to obtain the weighted trend deviation value; finally, these two weighted values ​​are added together to obtain the comprehensive anomaly index of the monitoring area. This index is a dimensionless value that integrates information from both local fluctuation anomalies and overall trend anomalies; the larger the value, the higher the probability of anomalies occurring in the area.

[0175] Furthermore, after determining the baseline value, each monitoring area is traversed, and its comprehensive anomaly index is compared with the baseline value. If the comprehensive anomaly index of an area is greater than the baseline value, the area is marked as an anomaly area of ​​historical buildings, and the area identifier of the anomaly area is recorded for subsequent detailed investigation and operation and maintenance management. All marked anomaly areas and their identifiers together constitute the anomaly area list.

[0176] In summary, by transforming complex time-series monitoring data into a quantized vector composed of four key morphological features, and by extracting local extreme points and monotonic variation intervals of the data sequence on the time axis, the intrinsic structure of data fluctuations can be accurately described. This makes it possible to mathematically compare fluctuation patterns in different time periods and monitoring areas, providing stable and quantifiable feature inputs for anomaly detection.

[0177] In summary, by using the mature metric method of Euclidean distance, the fluctuation patterns of current monitoring data and historical normal data are quantitatively compared. By calculating the straight-line distance between two four-dimensional vectors in space, the similarity of the patterns is transformed into a single numerical index. The larger the distance, the more significant the difference between the current fluctuation pattern and the typical historical pattern, thus providing an objective basis for judging whether there are local abnormal fluctuations.

[0178] In summary, by comparing the slope difference between the current overall trend of data and the historical long-term trend, we can effectively capture the cumulative changes of monitoring indicators over a long time scale, such as the slow evolution process of continuous structural deformation and gradual degradation of material properties. This deviation indicator complements the morphological difference value that reflects local fluctuations, together forming a comprehensive description of abnormal states.

[0179] In summary, by comprehensively considering both the morphological differences reflecting short-term local fluctuations and the deviation from the evolution of long-term overall trends, a single index that can comprehensively characterize the abnormal state of the monitored area is generated through a weighted fusion method. This index includes information on both sudden anomalies and gradual degradation, avoiding misjudgments or omissions that may be caused by a single indicator, and improving the accuracy and robustness of anomaly detection.

[0180] In summary, the method of dynamically determining the anomaly threshold based on the statistical characteristics of the data itself avoids the subjectivity and arbitrariness of manually setting the threshold. By analyzing the distribution of the comprehensive anomaly index of all monitored areas, it can adaptively identify those areas that deviate significantly from the normal range and record them together with the corresponding area identifier in the anomaly area list, providing accurate target positioning for subsequent targeted investigation and maintenance.

[0181] In this embodiment of the invention, the calculation formula for the comprehensive anomaly index is specifically used for:

[0182]

[0183] in, The comprehensive anomaly index, The weighting coefficient for the volatility pattern term. It is a natural constant. This represents the difference in fluctuation patterns. The mean of the fluctuation pattern difference. This represents the maximum difference in fluctuation patterns. It is a non-linear index with fluctuating patterns. The weighting coefficient for the trend evolution term. The deviation from the trend evolution, The standard deviation of the trend evolution. This represents the maximum deviation of the trend evolution. It is a non-linear index for trend evolution. For the weighting coefficients of the value coupling term, For comprehensive value coefficient, This represents the maximum value of the comprehensive value coefficient.

[0184] Specifically, the source of the fluctuation pattern difference value is the Euclidean distance calculation between the real-time fluctuation feature vector and the historical fluctuation feature vector. The calculation involves subtracting the number of rising segments in the real-time data from the number of rising segments in the historical data to obtain the first difference; subtracting the number of falling segments in the real-time data from the number of falling segments in the historical data to obtain the second difference; subtracting the number of peaks in the real-time data from the number of peaks in the historical data to obtain the third difference; and subtracting the number of troughs in the real-time data from the number of troughs in the historical data to obtain the fourth difference. These four differences are squared, summed, and then the square root of the sum is taken to obtain the final fluctuation pattern difference value. The source of the mean fluctuation pattern difference value is the arithmetic mean of the fluctuation pattern difference values ​​across all monitoring areas. The sum of the fluctuation pattern difference values ​​for each monitoring area is obtained, and this sum is divided by the total number of monitoring areas to obtain the mean fluctuation pattern difference value. The source of the maximum fluctuation pattern difference value is the comparison of the fluctuation pattern difference values ​​across all monitoring areas, selecting the largest value as the maximum fluctuation pattern difference value. The fluctuation pattern nonlinearity index is derived from a fixed value pre-set based on the structural characteristics of historical buildings and the type of monitoring indicators. This value is used to adjust the degree of nonlinear influence of fluctuation pattern differences in the comprehensive anomaly index. The trend evolution deviation is derived by performing linear regression on the overall trend of real-time data to obtain the real-time linear regression slope, and performing linear regression on the overall trend of historical archive data to obtain the historical linear regression slope. Subtracting the historical linear regression slope from the real-time linear regression slope yields the trend evolution deviation. The standard deviation of the trend evolution deviation is calculated by taking the standard deviation of the trend evolution deviation across all monitoring areas. First, the average trend evolution deviation of all monitoring areas is calculated. Then, the difference between the trend evolution deviation of each monitoring area and this average is calculated. Each difference is squared, and the sum of these squares is obtained. This sum is divided by the total number of monitoring areas, and the square root of the result is taken to obtain the standard deviation of the trend evolution deviation. The maximum value of the trend evolution deviation is obtained by comparing the trend evolution deviations of all monitoring areas and selecting the largest value as the maximum trend evolution deviation. The trend evolution nonlinearity index is derived from a pre-set fixed value based on the structural characteristics of historical buildings and the type of monitoring indicators. This value is used to adjust for the nonlinear influence of trend evolution deviation in the comprehensive anomaly index. The comprehensive value coefficient is a numerical value derived from a comprehensive assessment of the historical, cultural, and artistic value of the components within the monitored area, as well as their importance in the building structure. This value is obtained by a weighted average of scores given by multiple experts according to a unified value assessment standard. The maximum comprehensive value coefficient is determined by comparing the comprehensive value coefficients of all monitored areas and selecting the highest value as the maximum comprehensive value coefficient.The weighting coefficient for the fluctuation pattern term is a pre-set fixed value based on the impact of fluctuation patterns on building safety, used to adjust the proportion of fluctuation pattern difference terms in the comprehensive anomaly index. The weighting coefficient for the trend evolution term is a pre-set fixed value based on the impact of trend evolution on building safety, used to adjust the proportion of trend evolution deviation terms in the comprehensive anomaly index. The weighting coefficient η for the value coupling term is a pre-set fixed value based on the priority of value elements in operation and maintenance management, used to adjust the proportion of value coupling terms in the comprehensive anomaly index. The natural constant is a mathematically fixed constant used to construct the exponential decay function.

[0185] Furthermore, the significance of the formula lies in constructing a comprehensive evaluation index that can fully reflect the abnormal state of the monitored area of ​​historical buildings. The first term in the formula compares the fluctuation pattern difference value with the mean fluctuation pattern difference, normalizes it using the maximum fluctuation pattern difference value, and then introduces a natural constant to construct an exponential function. This means that when the fluctuation pattern difference value is less than the mean fluctuation pattern difference value, the output value of this term is compressed; when the fluctuation pattern difference value is greater than the mean fluctuation pattern difference value, the output value of this term is significantly amplified. The nonlinear exponent of fluctuation pattern further strengthens this amplification effect, thereby achieving sensitive capture of local fluctuation anomalies. The second term in the formula compares the trend evolution deviation degree with the standard deviation of the trend evolution deviation, normalizes it using the maximum trend evolution deviation degree, and then introduces a natural logarithmic function to construct a growth function. This means that when the trend evolution deviation degree is small, the output value of this term increases rapidly with the increase; when the trend evolution deviation degree is large, the growth rate of the output value of this term gradually slows down. The nonlinear exponent of trend evolution further adjusts the shape of this growth curve, thereby achieving a reasonable measurement of long-term trend deviation. The third term in the formula couples the comprehensive value coefficient with the fluctuation pattern difference value and the trend evolution deviation value in a ternary manner. It then performs global normalization by dividing by the product of the maximum comprehensive value coefficient, the maximum fluctuation pattern difference value, and the maximum trend evolution deviation value. This ensures that the higher the value of the monitoring area, the greater the weight of its fluctuation anomalies and trend anomalies in the comprehensive anomaly index, reflecting the greater impact of anomalies in high-value areas on the overall assessment. The three terms are multiplied by their respective weighting coefficients and then summed, allowing the comprehensive anomaly index to simultaneously reflect information from local fluctuation anomalies, long-term trend deviations, and the coupling of value elements.

[0186] In summary, the formula shows that as the difference in volatility patterns increases, the first output value gradually increases from zero. When the difference in volatility patterns exceeds the average difference, the growth rate accelerates significantly, eventually approaching the limit determined by the weighting coefficient of the volatility pattern term and the nonlinear exponent of the volatility pattern. As the deviation from the trend increases, the second output value increases rapidly from zero. As the deviation from the trend continues to increase, the growth rate gradually slows down, eventually approaching the limit determined by the weighting coefficient of the trend evolution term and the nonlinear exponent of the trend evolution. The third output value is proportional to the product of the comprehensive value coefficient, the difference in volatility patterns, and the deviation from the trend evolution (T). When all three factors increase simultaneously, this output value exhibits a cubic growth trend; however, when any one of these parameters is zero, the output value is zero. The overall trend of the comprehensive anomaly index is that it increases with the increase of the fluctuation pattern difference value, the trend evolution deviation, and the comprehensive value coefficient. When the fluctuation pattern difference value and the trend evolution deviation exceed the mean of fluctuation pattern difference and the standard deviation of trend evolution deviation, respectively, the growth rate of the comprehensive anomaly index will increase significantly. When the comprehensive value coefficient is close to the maximum value of the comprehensive value coefficient, the contribution of the third item to the comprehensive anomaly index will increase significantly, making the comprehensive anomaly index in high-value areas more sensitive to fluctuations and trend changes.

[0187] In this embodiment of the invention, when updating the dynamic maintenance file using the region identifier in the abnormal region list as an index to obtain the current file of the historical building, and matching the set of disease records in the current file with the historical maintenance records of the dynamic maintenance file to obtain the maintenance management plan for the historical building, the specific steps are as follows:

[0188] Based on the region identifier in the abnormal region list, locate the corresponding monitoring region in the dynamic operation and maintenance file, and lock the component node storage space contained in the monitoring region;

[0189] Create a special exploration data field in the component node storage space, and write the start and end identifiers of the abnormal time periods in the abnormal area list into the metadata area of ​​the special exploration data field;

[0190] Obtain a special survey data package for the component nodes during the abnormal period, the special survey data package including the physicochemical analysis results of material samples and the path of high-definition image files;

[0191] The feature data in the special survey data package is written into the corresponding subfields of the special survey data field to complete the data update of the component node and obtain the current archive of the historical building.

[0192] The disease type labels and disease severity levels in the existing disease record set are matched with the historical maintenance records in the dynamic operation and maintenance archive, and the matched historical maintenance measures are associated with the corresponding component node information to obtain the operation and maintenance management plan for the historical building.

[0193] Specifically, upon receiving the list of abnormal areas, the system uses a hash mapping method to precisely retrieve the storage location of the monitoring area corresponding to each area identifier recorded in the list within the index structure of the dynamic operation and maintenance file. The dynamic operation and maintenance file adopts a component-oriented storage model, where each monitoring area is associated with a specific set of component node storage spaces. These storage spaces centrally store the geometric data, attribute data, and historical monitoring data of all components within that area.

[0194] Specifically, within the locked component node storage space, the data structure of each component node is expanded by creating a new field specifically for storing specialized exploration data, named the Specialized Exploration Data Field. This field is designed as a composite data structure containing multiple sub-fields to accommodate different types of exploration results.

[0195] Specifically, based on the area identifiers and start and end dates of the abnormal periods in the list of abnormal areas, professional personnel or automated testing equipment are organized to conduct a special investigation of the status of component nodes within the designated abnormal areas during the abnormal periods. During the investigation, material samples are collected from each component node and sent to the laboratory for physicochemical analysis to obtain the physicochemical analysis results of the material samples.

[0196] Specifically, the special exploration data package of each component node is opened, the physicochemical analysis results of the material samples are extracted, and the results are written into the corresponding subfields of the newly created special exploration data fields in the storage space of that component node according to the predefined subfield names.

[0197] Specifically, from the updated existing archives, a set of defect records recorded in the special survey data fields of all abnormal area component node storage spaces is extracted. This set of records contains defect type tags for each component node. These defect type tags are used as keywords to perform precise matching searches in the historical maintenance record database of the dynamic operation and maintenance archives, finding all historical maintenance records with the same defect type tags, thus forming a preliminary matching record set.

[0198] Furthermore, after locating the monitoring area, the storage space of all component nodes contained in that area is locked through the row-level locking mechanism of the database to prevent concurrent write conflicts during subsequent data updates, ensuring data consistency and integrity. At this point, the location and locking operation of the component node storage space in the abnormal area is completed.

[0199] Furthermore, the start and end time markers corresponding to the abnormal period are extracted from the list of abnormal areas. These markers include the precise start and end times of the abnormality. These two time points are then written into the metadata area of ​​the special exploration data field as attribute information for that field, used to mark the specific time period targeted by this special exploration.

[0200] Furthermore, high-definition digital cameras or industrial endoscopes are used to photograph the surface condition and details of defects at component nodes, acquiring high-definition image files. These files are named according to the component node number and the shooting time, and stored in a designated file server, generating high-definition image file paths. These physicochemical analysis results and image file paths are then packaged according to component nodes to form a specialized investigation data package for each component node.

[0201] Furthermore, the high-resolution image file paths are written to the dedicated image path subfield within the specialized survey data field. During the writing process, it is ensured that the data format strictly matches the field definition; for example, physicochemical analysis results are stored in text or numerical form, while image paths are stored as strings. After completing the data writing for all component nodes within the abnormal areas, the previously locked storage space is released. At this point, the latest specialized survey data is added to the dynamic maintenance archive, and the entire archive is updated to its latest state, containing this new data, thus obtaining the current archive of the historical building.

[0202] Furthermore, historical maintenance records that are closest to the current disease severity level are selected from the preliminary matching record set, and the semantic similarity of the disease severity levels is compared. The historical maintenance measures text and maintenance effect evaluation data in the selected best matching records are associated with the component node information corresponding to the current disease record set. All association results are summarized and organized to form an operation and maintenance management plan for the current disease situation.

[0203] In summary, by using the area identifier output during the anomaly detection phase as the entry point, the system precisely navigates to the physical location in the dynamic operation and maintenance archive where all component data for that area are stored. A locking mechanism ensures that no concurrent access conflicts occur during subsequent data writing, providing accurate target positioning and a secure operating environment for the smooth storage of special exploration data, and guaranteeing the atomicity and consistency of the data update process.

[0204] In summary, the existing data structure is dynamically expanded to pre-allocate a dedicated storage area for the specific exploration results to be written. At the same time, the start and end times of the abnormal period are attached to this field as metadata, so that the exploration data stored later can clearly correspond to its target time window. This provides the necessary context information for distinguishing exploration records from different periods and conducting time-dimensional traceability analysis.

[0205] In summary, the results of the on-site investigation were structured and packaged. The physicochemical analysis results of the material samples provided quantitative evidence of the causes of the defects and the condition of the materials. The high-definition image file paths recorded intuitive evidence of the appearance characteristics of the defects. Together, they constituted a complete basis for assessing the health status of the components, providing real and verifiable data support for subsequent defect diagnosis and maintenance decisions.

[0206] In summary, by permanently linking and storing the evidence of defects obtained from on-site investigations with the corresponding components in the digital twin model, the previously scattered physical and chemical analysis reports and image files can be centrally managed according to component nodes. This completes the business loop from anomaly warning to on-site verification and data archiving. The updated current archives contain the latest defect information and have become an authoritative data source reflecting the current true condition of the building.

[0207] In summary, the knowledge value of the historical maintenance experience base is activated. By matching the types and severity levels of defects, successful cases of handling similar problems can be accurately retrieved from a vast historical record. The specific maintenance measures in these cases are then associated with the current defective components, allowing historical experience to be reused in new defect scenarios. The resulting operation and maintenance management solution provides managers with practically validated and highly targeted maintenance decision-making references.

[0208] In this embodiment of the invention, the step of matching the disease type tags and disease severity levels of the disease record set in the current archive with the historical maintenance records in the dynamic operation and maintenance archive, and then associating and summarizing the matched historical maintenance measures with the corresponding component node information to obtain the operation and maintenance management plan for the historical building, is specifically used for:

[0209] Retrieve historical maintenance records with the same disease type tag from the historical maintenance record database of the dynamic operation and maintenance archive to obtain a preliminary matching record set of the historical building;

[0210] The optimal matching record for the historical building is obtained by selecting the historical maintenance record with the smallest difference from the severity level of the disease from the preliminary matching record set.

[0211] The maintenance measures text and maintenance effect evaluation data in the optimal matching record are associated with the component node information corresponding to the disease record set to obtain the maintenance management plan for the historical building.

[0212] Specifically, in the historical maintenance record database of dynamic operation and maintenance archives, each disease type tag extracted from the current archive's disease record set is used as a search keyword to perform an exact match search on the historical maintenance record database. The historical maintenance record database is stored using a relational database, and each record contains a disease type tag field, the value of which is a standardized disease name extracted from the historical maintenance archive. During the search, the equality query command in the structured query language is executed to filter out all records in the database whose disease type tag field values ​​are exactly the same as the current search keyword.

[0213] Specifically, from the initial set of matching records, the historical maintenance record with the smallest difference from the current disease severity level is selected as the optimal matching record. First, the disease severity level needs to be quantified or mapped to continuous values ​​based on a more detailed level classification. The same quantification is performed on each disease severity level in the current disease record set.

[0214] Specifically, the optimal matching record is extracted completely from the historical maintenance record database, which includes maintenance measure text and maintenance effect evaluation data. At the same time, the corresponding component node information that needs maintenance is obtained from the defect record set, including the unique identifier of the component node, the spatial location of the node in the digital substrate, and the material properties of the node.

[0215] Furthermore, these selected records contain detailed information on various maintenance measures, maintenance times, maintenance personnel, and maintenance effect evaluations taken in the past for this type of disease. By compiling all the records that meet the criteria, a preliminary matching record set for historical buildings is obtained, which reflects all the historical experience in dealing with the same type of disease.

[0216] Furthermore, each historical maintenance record in the initial matching record set is traversed, and its quantified value of disease severity level is compared with the current quantified value of disease severity level. The absolute value of the difference between the two is calculated, representing the magnitude of the difference in severity levels. All calculated absolute difference values ​​are sorted, and the historical maintenance record corresponding to the minimum value is identified. If multiple records have the same minimum difference, their maintenance times are further compared, and the record with the most recent maintenance time is selected as the final optimal matching record. This ensures that the selected historical experience matches both the disease type and severity level, and is timely.

[0217] Furthermore, these maintenance measure texts and maintenance effect evaluation data are associated with the current component node information. For example, in the data structure of the operation and maintenance management plan, a plan entry is created for each component node, containing component node information and the corresponding historical best matching record. The plan entries for all component nodes are summarized to form a complete operation and maintenance management plan. This plan provides recommended maintenance measures for each abnormal component based on historical successful experience and tailored to the current type and severity of the defect.

[0218] In summary, it activates the knowledge reuse value of the historical maintenance experience base. Through precise matching of disease type tags, it quickly filters out all treatment records for the same type of disease from the maintenance archives accumulated over the years. It brings together the originally scattered and isolated historical data into a set of experiences directly related to the current problem, providing a rich reference case library for subsequent solution formulation. This avoids the blindness of formulating maintenance measures from scratch and significantly improves decision-making efficiency.

[0219] In summary, precise targeting of experience matching was achieved. By further comparing the severity levels of diseases based on the same disease type, the treatment cases in history that are closest to the current severity of diseases were selected. This process ensured the applicability and relevance of the selected historical experience, avoiding the risk of maintenance measures failing or over-maintenance due to large differences in disease severity, and enabling historical experience to serve the current problem in the most appropriate way.

[0220] In summary, the project successfully transformed historical experience into current decision-making, binding proven maintenance measures and quantifiable performance evaluations to specific damaged components. This resulted in a customized operation and maintenance plan that incorporates successful historical practices while precisely targeting the current location of the damage. This plan provides the on-site maintenance team with clear operational guidelines and expected results, enabling the effective inheritance and precise application of historical building maintenance experience.

[0221] Compared with the prior art, the present invention has the following beneficial effects:

[0222] 1. This digital twin-based historical building operation and maintenance management method constructs a digital foundation integrating geometric structure and historical information, and divides monitoring areas into multiple dimensions based on load-bearing structure, value elements, and functional spaces, achieving refined organization and spatial indexing of historical building data. This method precisely maps real-time operational data to the storage space of corresponding component nodes according to spatial location, forming dynamically updated operation and maintenance archives. Furthermore, it generates spatiotemporal distribution maps through visual coding technology, allowing for a direct presentation of the evolution patterns of monitoring indicators in both time and space dimensions. This significantly improves the management efficiency and readability of massive amounts of heterogeneous data, providing a highly structured and visualized data foundation for subsequent analysis.

[0223] 2. This method, through quantitative analysis of the differences in fluctuation patterns and trend deviations between real-time data and historical archive data in the spatiotemporal distribution map, can accurately identify abnormal areas and their occurrence periods, and bind abnormal information with area identifiers to form an abnormal area list. Based on this, the archives are dynamically updated and matched with historical maintenance records using abnormal areas as indexes, enabling the rapid generation of operation and maintenance management plans tailored to the current type and severity of damage. This achieves a data-driven process from anomaly early warning and on-site investigation to plan formulation, significantly improving the accuracy of anomaly response and the scientific nature of maintenance decisions, ensuring the effectiveness and relevance of historical building protection measures.

[0224] like Figure 2 The diagram shown is a functional block diagram of a historical building operation and maintenance management system based on digital twins provided in an embodiment of the present invention.

[0225] The historical building operation and maintenance management system 100 based on digital twins described in this invention can be installed in an electronic device. Depending on the functions implemented, the historical building operation and maintenance management system 100 based on digital twins may include a digital base module 101, a monitoring area module 102, an operation and maintenance file module 103, a spatiotemporal distribution map module 104, an abnormal area module 105, and an operation and maintenance management scheme 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 fixed functions, and are stored in the memory of the electronic device.

[0226] In this embodiment, the functions of each module / unit are as follows:

[0227] The digital infrastructure module constructs a digital infrastructure for the historical building based on historical data and point cloud data of the historical building.

[0228] The monitoring area module divides the monitoring area in the digital base according to the load-bearing structure, value element distribution and functional space of the historical building, and associates the monitoring area with data identifiers.

[0229] The operation and maintenance file module maps the operation data of the historical building to the storage space of the corresponding component in the digital base based on the monitoring area to which the operation data belongs, so as to obtain the dynamic operation and maintenance file of the historical building and the corresponding monitoring data stream.

[0230] The spatiotemporal distribution mapping module uses the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis to visually encode the monitoring data stream, thereby obtaining the spatiotemporal distribution map of the historical building.

[0231] The abnormal area module determines the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend evolution deviations between the real-time data and historical archive data in the spatiotemporal distribution map, and binds the area identifier and abnormal time period of the abnormal areas to obtain a list of abnormal areas of the historical buildings.

[0232] The operation and maintenance management solution module updates the dynamic operation and maintenance file using the area identifier in the abnormal area list as an index to obtain the current file of the historical building, and matches the set of disease records in the current file with the historical maintenance records of the dynamic operation and maintenance file to obtain the operation and maintenance management solution for the historical building.

[0233] 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.

[0234] 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.

[0235] 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.

[0236] 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.

[0237] The embodiments of this application 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.

[0238] 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 the operation and maintenance management of historical buildings based on digital twins, characterized in that, The method includes: A digital base for the historical building is constructed based on historical data and point cloud data of the historical building. Based on the load-bearing structure, distribution of value elements, and functional space of the historical building, monitoring areas are divided in the digital base, and data identifiers are associated with the monitoring areas. Based on the monitoring areas to which the operation data of the historical buildings belong, the operation data is mapped to the storage space of the corresponding components in the digital base to obtain the dynamic operation and maintenance files of the historical buildings and the corresponding monitoring data streams. Using the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis, the monitoring data stream is visually encoded to obtain the spatiotemporal distribution map of the historical buildings; Based on the difference in fluctuation patterns and trend deviation between the real-time data and historical archive data in the spatiotemporal distribution map, the abnormal areas of the historical buildings are determined, and the area identifiers and abnormal time periods of the abnormal areas are bound together to obtain a list of abnormal areas of the historical buildings. The dynamic maintenance file is updated using the area identifier in the abnormal area list as an index to obtain the current file of the historical building. The set of disease records in the current file is then matched with the historical maintenance records of the dynamic maintenance file to obtain the maintenance management plan for the historical building.

2. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 1, characterized in that, The construction of the digital base for the historical building based on historical data and point cloud data includes: The architectural drawings in the historical data are digitally scanned to obtain raster images of the historical buildings, and the raster images are geometrically corrected to obtain the base map of the historical buildings. Spatially register the point cloud data with the base map to obtain the alignment point set of the historical building; The building's outer contour line and edge feature line, which are concentrated at the alignment points, are connected according to their spatial adjacency to obtain the closed contour line of the historical building. The closed contour lines are then combined according to their spatial position within the historical building to obtain the wireframe structure of the historical building. The repair records and material information in the historical data are used as attribute fields and attached to the corresponding component nodes in the wireframe structure to obtain the digital base of the historical building.

3. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 1, characterized in that, The monitoring area is divided in the digital base according to the load-bearing structure, value element distribution, and functional space of the historical building, and a data identifier is associated with the monitoring area, including: The continuous component nodes belonging to the same structural force path in the digital base are grouped into the load-bearing structural area of ​​the historical building; The component nodes that are spatially adjacent and have the same value level in the digital base are merged into the value element area of ​​the historical building. Based on the functional zoning boundary lines in the digital base, component nodes located within the same functional zoning are aggregated into the functional space areas of the historical building; The boundaries of the load-bearing structure area, the value element area, and the functional space area are superimposed to obtain the comprehensive regional framework of the historical building. The overlapping part of the comprehensive regional framework is then cut and divided to obtain the defining blocks of the historical building. A feature region code is generated for the defined block, and the feature region code is written into the metadata field contained in the spatial block within the defined block to obtain the data identifier of the spatial block.

4. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 1, characterized in that, The monitoring areas to which the operational data of the historical buildings belong are respectively mapped to the storage space of the corresponding components in the digital substrate, thereby obtaining the dynamic operation and maintenance files of the historical buildings and the corresponding monitoring data streams, including: Obtain the operational data package of the historical building, wherein the operational data includes the collection timestamp, spatial coordinates of the collection point, and monitoring index values; The spatial coordinates of the acquisition points are matched with the spatial range of the monitoring area in the digital base to determine the reference monitoring area to which the running data packet belongs; The corresponding component node space in the digital substrate is located based on the data identifier of the benchmark monitoring area; The monitoring index values ​​are written sequentially into the monitoring data field of the component node space according to the timestamp order to obtain the independent area monitoring sequence of the historical building; By merging the monitoring sequences of the independent areas, the monitoring data stream of the historical building is obtained; The monitoring data stream and the real-time data fields of the corresponding component nodes are summarized to obtain the dynamic operation and maintenance file of the historical building.

5. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 1, characterized in that, The process of visually encoding the monitoring data stream, with the acquisition time as the horizontal axis and the spatial topology of the digital substrate as the vertical axis, to obtain the spatiotemporal distribution map of the historical buildings includes: Determine the mapping rules between indicator values ​​and visual feature parameters in the monitoring data stream. The mapping rules include: determining color hue based on the value range in the monitoring data stream, determining color saturation based on the rate of change of values ​​in the monitoring data stream, and determining graphic transparency based on the cumulative value in the monitoring data stream. Based on the area identifier of the monitoring area to which the monitoring data stream belongs, the corresponding component node in the digital base is retrieved, and the vertical axis coordinate value of the component node in the spatial topology of the digital base is read. The index values ​​are converted into corresponding visual feature parameter values ​​according to the mapping rules; A two-dimensional coordinate system for the historical building is constructed using the collection timestamp of the monitoring data stream as the horizontal axis and the vertical axis coordinate value as the vertical axis positioning point. Pixels are drawn point by point in the two-dimensional coordinate system using the visual feature parameter value as the pixel attribute value. The pixels are arranged into a regular row and column structure according to the time order of the horizontal axis and the spatial order of the vertical axis to obtain the spatiotemporal two-dimensional grayscale matrix of the historical building. The spatiotemporal two-dimensional grayscale matrix is ​​spliced ​​according to the spatial topology of the digital substrate to obtain the spatiotemporal distribution map of the historical building.

6. The historical building operation and maintenance management method based on digital twin as described in claim 1, characterized in that, The method involves determining the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend deviations between real-time data and historical archive data in the spatiotemporal distribution map, and binding the area identifiers and abnormal time periods of the abnormal areas to obtain a list of abnormal areas of the historical buildings, including: The real-time data and historical archive data of the spatiotemporal distribution map are decomposed into fluctuation patterns to obtain the number of rising segments, falling segments, peaks and troughs in the real-time data and the historical archive data, and to generate the real-time fluctuation feature vector and the historical fluctuation feature vector of the historical building. The Euclidean distance between the real-time fluctuation feature vector and the historical fluctuation feature vector is used as the fluctuation pattern difference value of the historical building. The difference between the linear regression slope of the real-time data and the linear regression slope of the historical archive data is used as the deviation of the trend evolution of the historical building. The comprehensive anomaly index of the monitored area is obtained by integrating the fluctuation pattern difference value with the trend evolution deviation. The statistical distribution characteristics of the comprehensive anomaly index are used to determine the anomaly judgment benchmark value of the historical building, and the monitoring area where the comprehensive anomaly index exceeds the anomaly judgment benchmark value is marked as the anomaly area of ​​the historical building.

7. The historical building operation and maintenance management method based on digital twin as described in claim 6, characterized in that, The formula for calculating the comprehensive anomaly index includes: in, The comprehensive anomaly index, The weighting coefficient for the volatility pattern term. It is a natural constant. This represents the difference in fluctuation patterns. The mean of the fluctuation pattern difference. This represents the maximum difference in fluctuation patterns. It is a non-linear index with fluctuating patterns. The weighting coefficient for the trend evolution term. The deviation from the trend evolution, The standard deviation of the trend evolution. This represents the maximum deviation of the trend evolution. It is a non-linear index for trend evolution. For the weighting coefficients of the value coupling term, For comprehensive value coefficient, This represents the maximum value of the comprehensive value coefficient.

8. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 1, characterized in that, The process involves updating the dynamic maintenance file using the region identifier in the abnormal region list as an index to obtain the current file of the historical building, and matching the set of disease records in the current file with the historical maintenance records of the dynamic maintenance file to obtain the maintenance management plan for the historical building, including: Based on the region identifier in the abnormal region list, locate the corresponding monitoring region in the dynamic operation and maintenance file, and lock the component node storage space contained in the monitoring region; Create a special exploration data field in the component node storage space, and write the start and end identifiers of the abnormal time periods in the abnormal area list into the metadata area of ​​the special exploration data field; Obtain a special survey data package for the component nodes during the abnormal period, the special survey data package including the physicochemical analysis results of material samples and the path of high-definition image files; The feature data in the special survey data package is written into the corresponding subfields of the special survey data field to complete the data update of the component node and obtain the current archive of the historical building. The disease type labels and disease severity levels in the existing disease record set are matched with the historical maintenance records in the dynamic operation and maintenance archive, and the matched historical maintenance measures are associated with the corresponding component node information to obtain the operation and maintenance management plan for the historical building.

9. The method for operation and maintenance management of historical buildings based on digital twins as described in claim 8, characterized in that, The process involves matching the disease type tags and disease severity levels in the current disease record set with the historical maintenance records in the dynamic operation and maintenance archive, and then associating and summarizing the matched historical maintenance measures with the corresponding component node information to obtain the operation and maintenance management plan for the historical building, including: Retrieve historical maintenance records with the same disease type tag from the historical maintenance record database of the dynamic operation and maintenance archive to obtain a preliminary matching record set of the historical building; The optimal matching record for the historical building is obtained by selecting the historical maintenance record with the smallest difference from the severity level of the disease from the preliminary matching record set. The maintenance measures text and maintenance effect evaluation data in the optimal matching record are associated with the component node information corresponding to the disease record set to obtain the maintenance management plan for the historical building.

10. A historical building operation and maintenance management system based on digital twins, used to implement the historical building operation and maintenance management method based on digital twins as described in any one of claims 1-9, characterized in that, The system includes: The digital infrastructure module constructs a digital infrastructure for the historical building based on historical data and point cloud data of the historical building. The monitoring area module divides the monitoring area in the digital base according to the load-bearing structure, value element distribution and functional space of the historical building, and associates the monitoring area with data identifiers. The operation and maintenance file module maps the operation data of the historical building to the storage space of the corresponding component in the digital base based on the monitoring area to which the operation data belongs, so as to obtain the dynamic operation and maintenance file of the historical building and the corresponding monitoring data stream. The spatiotemporal distribution mapping module uses the acquisition time of the monitoring data stream as the horizontal axis and the spatial topology of the digital substrate as the vertical axis to visually encode the monitoring data stream, thereby obtaining the spatiotemporal distribution map of the historical building. The abnormal area module determines the abnormal areas of the historical buildings based on the differences in fluctuation patterns and trend deviations between the real-time data and historical archive data in the spatiotemporal distribution map, and binds the area identifier and abnormal time period of the abnormal areas to obtain a list of abnormal areas of the historical buildings. The operation and maintenance management solution module updates the dynamic operation and maintenance file using the area identifier in the abnormal area list as an index to obtain the current file of the historical building, and matches the set of disease records in the current file with the historical maintenance records of the dynamic operation and maintenance file to obtain the operation and maintenance management solution for the historical building.

Images