A method, system and device for processing multi-source heterogeneous underground pipe network data

Abstract

The application provides a multi-source heterogeneous underground pipe network data processing method, system and device, wherein the method comprises the following steps: based on a preset underground pipe network semantic model and a preset pipe network constraint condition, performing alignment and connection processing on a plurality of source data sets of an obtained underground pipe network, and determining a candidate connection graph of pipe network objects in the underground pipe network; each vertex in the candidate connection graph represents a candidate fusion object, and a candidate connection edge represents a candidate connection relationship between two candidate fusion objects; based on a local connection cost of the candidate connection graph of any candidate pipe network object in the candidate fusion object under a target state, performing iterative update processing on the candidate fusion object and the candidate connection edge; based on candidate score values of the candidate pipe network objects in the updated candidate fusion object and the updated candidate connection edge obtained after the update, a target information record unit is determined. The scheme of the application can improve the fusion efficiency and accuracy of multi-source heterogeneous underground pipe network data, and reduce merging errors and unnecessary manual review.

Description

Technical Field

[0001] This invention relates to the field of underground pipeline network data processing technology, and in particular to a method, system and equipment for processing multi-source heterogeneous underground pipeline network data. Background Technology

[0002] Urban underground pipe network data typically originates from CAD as-built drawings, GIS pipe layers, BIM component models, geophysical survey results, inspection records, IoT sensor data, and historical ledgers. Due to significant differences in coordinate benchmarks, field naming, accuracy levels, update cycles, and topological representation methods among these sources, and the fact that existing platforms often rely on manual rules or simple field mapping for data entry, it is difficult to accurately handle issues such as homonyms, synonyms, coordinate offsets, endpoint breaks, overlapping pipe sections, missing manhole connection relationships, and attribute conflicts for the same pipe entity across multiple data sources. This results in low reliability, high maintenance costs, and difficulty in continuous updates for single-source maps.

[0003] In real-world projects, the more challenging issue isn't simply stacking layers together, but rather determining which records correspond to the same pipe segment, which endpoints should be connected, which attributes should be primary values, and which anomalies require manual review. Therefore, a fusion approach is needed that integrates semantic alignment, spatial registration, topology correction, and conflict resolution into a single processing chain. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method, system and equipment for processing multi-source heterogeneous underground pipeline network data, so as to improve the efficiency and accuracy of multi-source heterogeneous underground pipeline network data fusion processing, and provide accurate data support for the planning, construction and maintenance of underground pipeline networks throughout their entire life cycle.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a method for processing multi-source heterogeneous underground pipeline network data, comprising:

[0006] Obtain multiple source datasets of underground pipe networks;

[0007] Based on a preset underground pipeline semantic model and preset pipeline constraints, the data information of multiple source datasets is aligned and connected to determine a candidate connection graph of pipeline objects in the underground pipeline network. Each vertex in the candidate connection graph represents a candidate fusion object, and the candidate fusion object contains multiple candidate pipeline objects representing the same entity pipeline object. The candidate connection edges in the candidate connection graph represent the candidate connection relationship between two candidate fusion objects.

[0008] Based on the local connection cost of the candidate connection graph of any candidate pipeline object in the candidate fusion object under the target state, the candidate fusion object and the candidate connection edge are iteratively updated to obtain the updated candidate fusion object and the updated candidate connection edge.

[0009] Based on the candidate scores of the candidate pipeline objects and the candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the physical pipeline object corresponding to each updated candidate fusion object. The target information recording unit includes the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationship of the physical pipeline object, and connection relationship to be reviewed for the physical pipeline object corresponding to the current updated candidate fusion object.

[0010] In one embodiment, based on a preset underground pipeline semantic model and preset pipeline constraints, the data information of multiple source datasets is aligned and connected to determine candidate connection graphs of pipeline objects in the underground pipeline network, including:

[0011] Based on the preset underground pipeline semantic model, the multiple source datasets are standardized to determine the standard information recording unit for each pipeline object in the underground pipeline network. The standard information recording unit includes the standard semantics of the current pipeline object, the standard semantic confidence, the data information representing the current pipeline object in each source dataset, and the source dataset identifier of the data information.

[0012] Based on the standard information recording unit of each pipeline object and the first preset pipeline constraint conditions, the candidate connection diagram of the pipeline objects in the underground pipeline network is determined.

[0013] In one embodiment, the multiple source datasets are standardized based on the preset underground pipeline semantic model to determine the standard information recording unit for each pipeline object in the underground pipeline network, including:

[0014] Based on multiple source datasets, an initial information recording unit is determined for each pipeline object in the underground pipeline network; the pipeline object includes underground pipe segments and pipeline facilities, and the initial information recording unit includes the original semantic record, data information, and source dataset identifier of the data information representing the pipeline object in each source dataset.

[0015] Based on the preset underground pipeline semantic model, the original semantic records corresponding to each source dataset in the initial information recording unit of each pipeline object are mapped to obtain the standard semantics and standard semantic confidence of each pipeline object, and a standard information recording unit for each pipeline object is generated; the standard semantic confidence represents the accuracy of mapping the original semantic records to standard semantics.

[0016] In one embodiment, based on the standard information recording unit of each pipeline object and the first preset pipeline constraint conditions, a candidate connection diagram of the pipeline objects in the underground pipeline network is determined, including:

[0017] Based on the standard information recording unit, the network objects contained in the multiple source datasets are divided to obtain multiple candidate identity graphs. The candidate identity graphs contain multiple candidate network objects representing the same entity, and each candidate network object serves as a vertex of the candidate identity graph. The connecting edges in the candidate identity graphs represent the candidate identity edges of the candidate network objects.

[0018] The candidate identity graph is used as a candidate fusion object, and based on the candidate fusion object and the first preset pipeline constraint, the candidate connection edge between the two candidate fusion objects is determined.

[0019] Based on the candidate fusion objects and the candidate connection edges, the candidate connection graph is constructed and obtained.

[0020] In one embodiment, based on the local connectivity cost of the candidate connection graph of any candidate network object in the candidate fusion objects under the target state, the candidate fusion objects and the candidate connection edges are iteratively updated to obtain updated candidate fusion objects and updated candidate connection edges, including:

[0021] Determine the first local connection cost of the current candidate connection graph of the current candidate pipeline object in the split state and the second local connection cost of the current candidate connection graph of the current candidate pipeline object in the fusion state;

[0022] Based on the first local connection cost and the second local connection cost of the current candidate connection graph, the current candidate fusion object and the candidate connection edges in the current candidate connection graph are updated to obtain the updated candidate fusion object of the current candidate fusion object and the updated candidate connection edges of the candidate connection edges in the current candidate connection graph.

[0023] The updated candidate fusion object is used as the current candidate fusion object for the next round of iterative update processing, and the determination and update process of local connection cost is repeated until the iteration number threshold or preset iteration condition is reached.

[0024] In one embodiment, determining the first local connectivity cost of the current candidate connection graph of the current candidate network object in the split state and the second local connectivity cost of the current candidate connection graph of the current candidate network object in the fusion state includes:

[0025] The cost of local connectivity is determined using the following formula:

[0026]

[0027] Where i=1,2 represents the target state, i=1 indicates that the current candidate pipeline object in the current candidate fusion object is in the split state, and i=2 indicates that the current candidate pipeline object in the current candidate fusion object is in the fusion state; Let the cost of local connectivity be denoted as , where the cost of the first local connectivity is denoted as . The cost of the second local connection is expressed as ; This indicates the conflict value between the two ends of the current candidate pipeline object; This indicates the diameter continuity deviation value between the two ends of the current candidate pipeline object; This indicates the elevation interval conflict value between the two ends of the current candidate pipeline object; This indicates the flow direction conflict value at the connection end of the current candidate pipeline object; This indicates an outlier value in the node degree of the current candidate connectivity graph under the target state; This indicates the discontinuity of the node neighborhood in the current candidate connectivity graph under the target state; These represent the weights of the corresponding parameters.

[0028] In one embodiment, based on the first local connection cost and the second local connection cost of the current candidate connection graph, the current candidate fusion object and the candidate connection edges in the current candidate connection graph are updated to obtain the updated candidate fusion object of the current candidate fusion object and the updated candidate connection edges of the candidate connection edges in the current candidate connection graph, including:

[0029] Determine the cost difference between the first local connection cost and the second local connection cost; in response to the cost difference being greater than or equal to a splitting threshold, determine the candidate identity edge of the current candidate pipeline object as an invalid identity edge.

[0030] Based on the invalid identity edge, the corresponding current candidate pipeline object in the current candidate fusion object is split to obtain the updated fusion object;

[0031] Based on the updated fusion object and the second preset pipeline constraint, the candidate connection edge for updating between the two updated fusion objects is determined.

[0032] In one embodiment, based on the candidate score values ​​of the candidate network objects and the candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the entity network object corresponding to each updated candidate fusion object, including:

[0033] Based on the standard information recording unit of each candidate pipeline object in the updated candidate fusion object, determine the attribute candidate score value and geometric candidate score value of each candidate pipeline object;

[0034] Based on the updated candidate connection edges, determine the connection candidate score value of the updated fusion object;

[0035] Based on the attribute candidate score, the geometric candidate score, and the connection candidate score, the multiple candidate pipeline objects of the updated candidate fusion object are sorted, and the primary attribute features, alternative attribute features, primary geometric features, alternative geometric features, connection relationships of physical pipeline objects, and connection relationships to be reviewed are determined according to the sorting order to generate the target information recording unit.

[0036] Embodiments of the present invention also provide a system for processing multi-source heterogeneous underground pipeline network data, comprising:

[0037] The acquisition module is used to acquire multiple source datasets of underground pipe networks;

[0038] The processing module is used to align and connect centralized data information from multiple source datasets based on a preset underground pipeline semantic model and preset pipeline constraints, to determine candidate connection graphs of pipeline objects in the underground pipeline network. Each vertex in the candidate connection graph represents a candidate fusion object, and the candidate fusion object contains multiple candidate pipeline objects representing the same entity. Candidate connection edges in the candidate connection graph represent candidate connection relationships between two candidate fusion objects. Based on the local connection cost of the candidate connection graph for any candidate pipeline object in the candidate fusion object under the target state, the module performs alignment and connection processing on the candidate fusion object and the candidate pipeline objects. The connection edges are iteratively updated to obtain updated candidate fusion objects and updated candidate connection edges. Based on the candidate network objects and candidate score values ​​of the updated candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the physical network object corresponding to each updated candidate fusion object. The target information recording unit contains the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationship of the physical network object, and connection relationship to be reviewed for the physical network object corresponding to the current updated candidate fusion object.

[0039] Embodiments of the present invention also provide a computing device, comprising: a memory for storing one or more programs; and one or more processors for executing the one or more programs to implement the method described above.

[0040] The above-described solution of the present invention has at least the following beneficial effects:

[0041] 1. The solution of this invention first performs semantic alignment of data in multi-source datasets by using a preset underground pipeline network semantic model, and then performs identity verification and connection relationship modeling of pipeline network objects based on the aligned data information, fundamentally avoiding erroneous merging of pipeline network objects; at the same time, it breaks down data barriers between planning, surveying, operation and maintenance departments, realizes true integration and sharing of multi-source data, and breaks down data information silos.

[0042] 2. After confirming the identity information and connection relationship of the pipeline network object, the correctness of the confirmed identity information is verified by back-checking the local connection cost of the candidate connection graph of the candidate pipeline network object in the target state. This enables the candidate fusion object formed by the pipeline network object to be updated, realizing bidirectional correction of the identity confirmation and recognition of the pipeline network object, and ensuring the accuracy of the subsequent generation of target information recording units.

[0043] 3. The target information record unit generated by this solution ultimately includes the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, corresponding physical pipeline object connection relationship and connection relationship to be verified, so as to maximize the use of the advantages of each source dataset and provide accurate data support for the refined management and operation and maintenance of urban underground pipeline networks.

[0044] It should be understood that the implementation of any embodiment of the present invention does not mean that it will simultaneously possess or achieve multiple or all of the above-mentioned beneficial effects. Attached Figure Description

[0045] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0046] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0047] Figure 1 This is a flowchart illustrating a method for processing multi-source heterogeneous underground pipeline network data provided in an embodiment of the present invention.

[0048] Figure 2 This is a candidate identity map formed by multiple candidate network objects, provided in an optional embodiment of the present invention;

[0049] Figure 3 This is a candidate connection diagram formed by multiple candidate fusion objects, provided in an optional embodiment of the present invention;

[0050] Figure 4 This is a schematic diagram illustrating the specific implementation process of iteratively updating candidate fusion objects and candidate connection edges according to an optional embodiment of the present invention;

[0051] Figure 5 This is a schematic block diagram of a multi-source heterogeneous underground pipeline network data processing system provided in an embodiment of the present invention;

[0052] Figure 6 This is a schematic block diagram of an electronic device provided in an embodiment of the present invention; and

[0053] Figure 7 This is a schematic block diagram of a computing device provided in an embodiment of the present invention. Detailed Implementation

[0054] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0055] It should be understood that the terms "comprising / including," "consisting of," or any other variations are intended to cover non-exclusive inclusion, such that a product, apparatus, process, or method that comprises a list of elements includes not only those elements but may also include, where necessary, other elements not expressly listed, or elements inherent to such a product, apparatus, process, or method. Without further limitation, an element defined by the phrases "comprising / including," "consisting of," does not exclude the presence of additional identical elements in the product, apparatus, process, or method that includes said element.

[0056] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0057] like Figure 1As shown, an embodiment of the present invention proposes a method for processing multi-source heterogeneous underground pipeline network data, including:

[0058] Step 11: Obtain multiple source datasets of the underground pipe network;

[0059] Step 12: Based on the preset underground pipeline semantic model and preset pipeline constraints, perform semantic alignment and connection processing on the data information in multiple source datasets to obtain a candidate connection graph of pipeline objects in the underground pipeline network; each vertex in the candidate connection graph represents a candidate fusion object, and the candidate fusion object contains multiple candidate pipeline objects representing the same entity; the candidate connection edges in the candidate connection graph represent the candidate connection relationship between two candidate fusion objects.

[0060] Step 13: Based on the local connection cost of the candidate connection graph of any candidate pipeline object in the candidate fusion object under the target state, perform iterative update processing on the candidate fusion object and the candidate connection edge to obtain the updated candidate fusion object and the updated candidate connection edge.

[0061] Step 14: Based on the candidate scores of the candidate pipeline objects and the candidate connection edges in the updated candidate fusion objects, determine the target information recording unit for the corresponding physical pipeline object of each updated candidate fusion object; the target information recording unit contains the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationship of the physical pipeline object and connection relationship to be reviewed for the physical pipeline object corresponding to the current updated candidate fusion object.

[0062] In this embodiment, multiple source datasets may include GIS pipeline layer sets, pipeline CAD as-built drawings sets, pipeline detection and inspection results sets, and pipeline BIM construction model sets; each dataset may include data information describing the same physical pipeline object, which may specifically include the chart or field name and attribute values ​​(such as coordinates, size parameters, geometric parameters, etc.) of the physical pipeline object, and the format and granularity of the data information describing the same physical pipeline object in each dataset are different;

[0063] Since the descriptions of the same entity pipeline object in the initially obtained source datasets are inconsistent, we can first align the data information in different source datasets using a pre-defined underground pipeline semantic model to confirm the identity representation (corresponding candidate pipeline object) of each same pipeline entity in different source datasets, and cluster them to form candidate fusion objects. Subsequently, based on pre-defined pipeline constraints, we connect the candidate fusion objects that identify different entity pipeline objects, thereby obtaining a candidate connection graph that may represent the underground pipeline topology. By first confirming the identity representation of pipeline objects in different source datasets and then establishing a candidate connection graph of the connection relationships between candidate fusion objects, we avoid the problem of the same entity object appearing repeatedly due to directly using a single pipeline object in the source dataset for relationship connection, thereby reducing the error rate of merging and connection, and making the constructed candidate connection graph more accurate.

[0064] To avoid errors in identity verification and grouping of the same entity network object, we can verify the identity representation of the same entity network object in different source datasets based on the local connectivity cost of any candidate network object in the candidate connection graph under the target state. Then, we can re-determine the candidate network object based on the verified identity to update the candidate network object. Here, the target state can include a split state and a fusion state. The split state represents separating any candidate network object from the current candidate network object (assuming the current candidate network object and the current candidate network object represent different entity network objects). Simultaneously, in the split state, we confirm the local connectivity cost of the current candidate network object in the candidate connection graph when it connects with its neighboring candidate network objects after splitting. The fusion state can be directly viewed as the current candidate network object being in the current candidate network object. Simultaneously, in the fusion state, we confirm the local connectivity cost of the current candidate network object in the candidate connection graph when it connects with its neighboring candidate network objects. Based on the local connectivity costs under different states, we can achieve bidirectional verification of the identity of the same entity network object, thereby improving the accuracy of identity judgment and candidate connection judgment.

[0065] In a specific implementable instance, there are three candidate pipeline objects in the current candidate fusion object E, denoted as E1, E2, and E3 respectively. The split state of any two candidate pipeline objects can be represented as ((E1, E2), E3), ((E3, E2), E1), and ((E1, E3), E2 respectively). After splitting, the current candidate pipeline object that is separated from the current candidate fusion object is regarded as a separate candidate fusion object, and the local connection cost is calculated simultaneously with the split current fusion object.

[0066] Furthermore, based on the candidate score values ​​of the data information of each candidate pipeline object in the updated candidate fusion object, as well as the candidate score values ​​of the updated candidate connection edges, multiple candidate pipeline objects and their corresponding data information in the updated candidate fusion object are filtered to obtain the optimal data record of the corresponding physical pipeline object, so as to maximize the use of the advantages of each source data. At the same time, the target information record unit of the corresponding physical pipeline object is formed based on the optimal data record, thereby providing accurate data support for the planning, construction, and maintenance of underground pipelines throughout their entire life cycle.

[0067] In an optional embodiment of the present invention, step 12 above may include:

[0068] Step 121: Based on the preset underground pipeline semantic model, standardize multiple source datasets to determine the standard information recording unit for each pipeline object in the underground pipeline network. The standard information recording unit includes the standard semantics of the current pipeline object, the standard semantic confidence, the data information representing the current pipeline object in each source dataset, and the source dataset identifier of the data information.

[0069] In this embodiment, the standard information recording unit represents the information recording geometry after standardizing the data information of the corresponding pipeline object represented in each source dataset; the standardization process facilitates the subsequent screening and classification of the same entity pipeline object based on multiple source datasets.

[0070] Specifically, step 121 above may include:

[0071] Step 1211: Based on multiple source datasets, determine the initial information recording unit for each pipeline object in the underground pipeline network; here, the pipeline object may include underground pipe sections and pipeline facilities, and the pipeline equipment may specifically include point-like or area-like pipeline ancillary facilities such as pipe wells, valves, inspection wells, rainwater inlets, and fire hydrants.

[0072] Step 1212: Based on the preset underground pipeline semantic model, the original semantic records corresponding to each source dataset in the initial information recording unit of each pipeline object are mapped to obtain the standard semantics and standard semantic confidence of each pipeline object, and a standard information recording unit for each pipeline object is generated; the standard semantic confidence represents the accuracy of mapping the original semantic records to the standard semantics.

[0073] In this embodiment, after obtaining multiple source datasets, the different source datasets can be parsed separately. Specifically, the spatial reference, layer mode, point, line and polygon features, attribute table and existing topological relationships of the GIS pipeline layer set are read, and the feature ID and layer path are retained. The layer name, line type, color, block reference, text annotation and polyline endpoints of the pipeline CAD as-built drawing set are parsed. The measurement points, trajectory, depth / elevation, equipment error, work batch, facility number, address description, pipe diameter, material, ownership and status of the pipeline detection and inspection results set are read. The component type, local / global coordinates, pipe diameter, material, port and component connection of the pipeline BIM construction model set are read.

[0074] Furthermore, the parsed data can be preprocessed. Preprocessing can specifically include coordinate system unification, geometric type conversion, field cleaning, unit conversion, null value identification, and initial screening of duplicate records. For example, the pipe diameter field can be standardized to millimeters, the burial depth field to meters, and the coordinates to the target engineering coordinate system. Simultaneously, the preprocessed data is packaged to generate initial information record units for each pipeline network object. These initial information record units are then labeled with source datasets to determine the information source of the current pipeline network object.

[0075] Here, the pre-defined semantic model of the underground pipeline network can be a database composed of an entity dictionary set, an attribute ontology set, and a relation schema set. The entity dictionary set defines unified semantic types and aliases for pipe segments, pipe wells, inspection wells, valves, pressure regulating facilities, rainwater inlets, and cable wells; the attribute ontology set defines standard attributes, units, and allowable value ranges for pipe diameter, material, burial depth, starting and ending elevations, pressure rating, ownership, and status; and the relation schema set defines relationships such as connection, subordination, load-bearing, intersection, proximity, upstream, and downstream.

[0076] Based on the preset underground pipeline network semantic model, the original semantic records in the data information of the initial information recording unit are mapped to obtain the standard semantic type (or standard attribute) of each pipeline object, which provides a basis for the subsequent classification of pipeline objects and the construction of candidate connection diagrams. It should be understood that the mapping process here is to unify the record name, parameter value, parameter unit, and association relationship of adjacent pipeline objects in the original semantic records to obtain their respective standard semantics.

[0077] Preferably, in an implementable embodiment of the present invention, the determination of whether the original semantic record undergoes standard semantic mapping processing can be based on the standard semantic confidence score of the original semantic record mapped to standard semantics; more preferably, the standard semantic confidence score can be expressed by the following formula:

[0078]

[0079] in, The standard semantic confidence level is represented by T; name similarity by A; value range distribution similarity by U; unit and encoding compatibility by R; and historical confirmation by H. These represent the weights of the corresponding parameters, which can be obtained through training or statistical calibration based on historical confirmed samples. When historical samples are insufficient, initial weights are set according to expert rules and iteratively updated based on subsequent manual review results. Each weight is normalized.

[0080] Here, after obtaining multiple source datasets, the GIS pipeline layers, CAD as-built drawings, inspection results, and BIM component models are first parsed, their coordinates unified, fields cleaned, units converted, and source identifiers recorded to form initial information recording units for each pipeline network object. From these initial information recording units, the following are extracted: T (name similarity), A (value range distribution similarity), U (unit and coding compatibility), R (similarity of association between neighboring pipeline network objects), and H (historical confirmation). Specifically, T (name similarity) is obtained by matching field names, layer names, component names, block names, text annotations, and entity dictionaries / attribute ontology; A (value range distribution similarity) is obtained by the value range, category frequency, and histogram distribution of the same attribute; U (unit and coding compatibility) is obtained by the compatibility of unit conversion results and coding rules; R (similarity of association between neighboring pipeline network objects) is obtained by the type, number of ports, and connection / subordination / upstream / downstream relationships of adjacent pipeline network objects; and H (historical confirmation) is obtained by statistical analysis of historical manual confirmations, historical fusion results, or historical rejection records. After confirming the standard semantic confidence of the corresponding pipeline network object, when its corresponding standard semantic confidence... When the output is not less than the automatic mapping threshold, the output target is a unified standard semantic (or standard attribute), that is, the initial semantic record in the current pipeline object is directly replaced with the standard semantic; when the standard semantic confidence level is... When the data is within the review threshold range, retain the initial information record unit of the current pipeline object and mark the initial information record unit of the current pipeline object as pending review; when the standard semantic confidence level is within the threshold range... If the value is less than the automatic mapping threshold, the initial information recording unit of the current pipeline object will be marked as unmatched.

[0081] In an optional embodiment of the present invention, step 12 above may include:

[0082] Step 122: Based on the standard information recording unit of each pipeline object and the first preset pipeline constraint conditions, determine the candidate connection diagram of the pipeline objects in the underground pipeline network.

[0083] In this embodiment, the first preset pipeline network constraints may include pipeline type or media compatibility constraints, endpoint spatial proximity and uncertainty domain intersection constraints, pipe diameter / material / pressure level continuity constraints, burial depth or elevation interval constraints, flow direction and upstream and downstream sequence constraints, facility capacity constraints, node degree and topological connectivity constraints, and multi-source record consistency constraints.

[0084] Specifically, step 122 above may include:

[0085] Step 1221: Based on the standard information recording unit, the network objects contained in multiple source datasets are divided to obtain multiple candidate identity graphs. Each candidate identity graph contains multiple candidate network objects representing the same entity, and each candidate network object serves as a vertex of the candidate identity graph. The connecting edges in the candidate identity graph represent the candidate identity edges of the candidate network objects.

[0086] Step 1222: Take the candidate identity graph as a candidate fusion object, and determine the candidate connection edge between the two candidate fusion objects based on the candidate fusion object and the first preset pipeline constraint conditions;

[0087] Step 1223: Based on the candidate fusion objects and candidate connection edges, construct and obtain the candidate connection graph.

[0088] In this embodiment, the standard semantic confidence level in the information recording unit can be used to first confirm whether any two network objects represent the same entity network object candidate identity edge; preferably, the candidate identity edge can be calculated using the following formula:

[0089]

[0090] in, This represents the candidate identity edge score between pipeline object m and pipeline object n; This represents the smaller of the two network objects m and n, indicating the standard semantic confidence. This represents the overlap rate of the geometric uncertainty domains of pipeline object m and pipeline object n (that is, the overlap ratio of the geometric uncertainty domains of the two objects after the coordinates of pipeline object m and pipeline object n are unified. The geometric uncertainty domain is obtained by correcting the geometric dimensions of the pipeline object in the CAD as-built drawing or BIM component model based on the coordinates of the pipeline object). This represents the weighted compatibility of comparable attributes between pipeline network object m and pipeline network object n (that is, the compatibility of comparable attributes such as pipe diameter, material, burial depth, elevation, ownership, and status between pipeline network object m and pipeline network object n after unit conversion). This represents the similarity of the association relationships between pipeline object m and pipeline object n with their neighboring pipeline objects (that is, the similarity of the set of neighboring pipeline objects, the number of ports, and the types of associated facilities between pipeline object m and pipeline object n). This indicates the support level in the historical information record unit that confirms that pipeline object m and pipeline object n are the same entity pipeline object; This represents the time conflict penalty value between pipeline object m and pipeline object n (that is, the conflict between their collection time, completion time, inspection time or version time). These represent the regression coefficients for each feature term, which are obtained based on training or statistical calibration using historically confirmed samples. When samples are insufficient, they can be preset according to expert rules and iteratively updated based on subsequent manually reviewed samples. The corresponding time conflict penalty term is indicated by a negative sign in the formula, signifying its negative effect.

[0091] When the candidate identity edge score of any two network objects is not less than the identity edge threshold (one identity edge threshold corresponds to one type of entity network object), the two network objects are confirmed to be the same entity network object. The standard information record units of all network objects are traversed, and the candidate identity edge score of any two network objects is calculated sequentially. All network objects are then divided, and multiple network objects belonging to the same entity network object are confirmed as candidate network objects. A candidate identity graph is constructed using the candidate network objects as vertices and the candidate identity edges between two candidate network objects as connecting edges, as detailed below. Figure 2 As shown in the figure, the standard information recording unit of candidate pipeline object E1 can be obtained based on the source dataset - GIS pipeline layer set, the standard information recording unit of candidate pipeline object E2 can be obtained based on the source dataset - pipeline CAD as-built drawing set, and the standard information recording unit of candidate pipeline object E3 can be obtained based on the source dataset - pipeline detection and inspection result set.

[0092] Furthermore, the candidate identity connection graph is treated as a whole as a candidate fusion object, which stores standard information record units of all candidate pipeline objects; it should be known that one candidate fusion object corresponds to one type of entity pipeline object.

[0093] Here, a candidate connection graph is established with candidate fusion objects as vertices. Based on the data information in the standard information recording unit and the first preset pipeline constraint conditions, it is determined whether there is a candidate connection edge representing a topological connection relationship between any two candidate fusion objects. Here, the existence of a candidate connection edge between two fusion objects can be determined by connectivity scoring. Preferably, the connectivity score can be determined by the following formula:

[0094]

[0095] in, The connectivity score between candidate fusion object u and candidate fusion object v; The endpoint support of candidate fusion object u and candidate fusion object v can be represented by the intersection rate of the uncertainty domains of the endpoints (the endpoints refer to the distance from the endpoints to the facility center or boundary and the directional angle). This indicates the facility support ratio between candidate fusion object u and candidate fusion object v (the consistency ratio between candidate fusion object u and candidate fusion object v in terms of well, valve, port, and ledger facility number; when the two belong to the same well, valve, port, or ledger facility number, they are considered to be consistent, and the consistency ratio = the number of times they are consistent / the total number of comparisons). This represents the physical constraints between candidate fusion object u and candidate fusion object v (the proportion of the data information of the physical network objects corresponding to candidate fusion object u and candidate fusion object v that meet the corresponding professional rules). The neighborhood continuity after connecting candidate fusion object u and candidate fusion object v is represented by the degree of the nodes after connection, the direction of entry and exit, the order of upstream and downstream, the adjacency type and the local topological continuity. This represents the multi-source support (the proportion of records in different source datasets where the entity network object corresponding to candidate fusion object u and the entity network object corresponding to candidate fusion object v are connected to each other). These represent the weights of the corresponding parameters. The corresponding weights can be obtained through historical connection confirmation sample statistics or training; when historical samples are insufficient, initial weights are preset according to the importance of endpoints, facilities, physical constraints, neighborhood continuity, and multi-source support, and then normalized. When the connectivity score is greater than or equal to the connectivity threshold, a candidate connection edge is determined between candidate fusion object u and candidate fusion object v. After traversing and calculating all candidate fusion objects, the corresponding candidate connectivity graph can be obtained, as shown in the figure below. Figure 3 As shown.

[0096] See Figure 4 In an optional embodiment of the present invention, step 13 above may include:

[0097] Step 131: Determine the first local connection cost of the current candidate network object in the split state and the second local connection cost of the current candidate network object in the fusion state.

[0098] Here, the current candidate pipeline object is in a fused state, that is, the current candidate pipeline object is in the current candidate fused object. The local connection cost calculated at this time is the second local connection cost when the current candidate fused object is connected to its neighboring candidate fused objects in the candidate connection graph. When the current candidate pipeline object is in a split state, that is, the current candidate pipeline object is separated from the current candidate fused object and treated as an independent connection individual, the local connection cost calculated when the connection individual and the split candidate fused object are connected to their neighboring candidate fused objects is the first local connection cost.

[0099] When a candidate pipeline object is merged into another candidate fusion object, it is based on candidate identity edge fusion. When a candidate pipeline object is separated from the current candidate fusion object, its candidate identity edges are also separated. Since there are multiple candidate identity edges of candidate pipeline objects in the candidate fusion object, before each iteration update, the multiple candidate identity edges are first arranged in ascending order, and the two candidate pipeline objects corresponding to the smallest candidate identity edge score are selected as the pipeline objects to be evaluated in the current iteration. The local connectivity cost of the candidate connection graph is then calculated when the pipeline objects to be evaluated are in the target state.

[0100] Preferably, the corresponding local connectivity cost can be determined using the following formula:

[0101]

[0102] Where i=1,2 represents the target state, i=1 indicates that the current candidate pipeline object in the current candidate fusion object is in the split state, and i=2 indicates that the current candidate pipeline object in the current candidate fusion object is in the fusion state; Let the cost of local connectivity be denoted as , where the cost of the first local connectivity is denoted as . The cost of the second local connection is expressed as ; This indicates the conflict value between the two ends of the current candidate pipeline network object (obtained from the pipeline type or media compatibility at both ends; if the pipeline types at both ends are the same or compatible, the value is 0; otherwise, it is 1). This indicates the difference in pipe diameter between the two ends of the current candidate pipeline network object; This represents the elevation interval conflict value between the two ends of the current candidate pipeline object (the value range is [0, 1]). This indicates the flow conflict value of the current candidate pipeline object's connection end (indicating the number of abnormal flow connection ends on the current candidate pipeline object when it is in a split or merge state; abnormal flow connection ends indicate that when the current candidate pipeline object is connected to other current candidate pipeline objects, there is only inflow (no outflow) or outflow (no inflow) between them). This indicates the abnormal node degree value in the current candidate connection graph under the target state (indicating the number of abnormal vertices in the current candidate connection graph when it is in a split or merge state, and the corresponding abnormal vertex indicates that the number of candidate connection edges on that vertex exceeds the preset number). This indicates the discontinuity of the node neighborhood in the current candidate connection graph under the target state (obtained by the number or proportion of discontinuous relationships between adjacent pipe segments, facilities, and ports within the neighborhood). These represent the weights of the corresponding parameters. The corresponding weights can be statistically calibrated based on historical split / retained samples. When historical samples are insufficient, initial weights are preset according to the importance of professional constraints such as pipeline type, pipe diameter, elevation, flow direction, node degree, and neighborhood discontinuity, and then normalized.

[0103] Continue reading Figure 4 In an optional embodiment of the present invention, step 13 above may include:

[0104] Step 132: Based on the first local connection cost and the second local connection cost of the current candidate connection graph, update the current candidate fusion object and the candidate connection edges in the current candidate connection graph to obtain the updated candidate fusion object of the current candidate fusion object and the updated candidate connection edges of the candidate connection edges in the current candidate connection graph.

[0105] Specifically, step 132 above may include:

[0106] Step 1321: Determine the cost difference between the first local connection cost and the second local connection cost. In response to the cost difference being greater than or equal to the splitting threshold, determine the candidate identity edge of the current candidate pipeline object as an invalid identity edge.

[0107] Step 1322: Based on the invalid identity edge, split the current candidate pipeline object corresponding to the current candidate fusion object to obtain the updated fusion object;

[0108] Step 1323: Based on the updated fusion object and the second preset pipeline constraint conditions, determine the candidate connection edge for updating between the two updated fusion objects.

[0109] Here, the cost difference Specifically, it can be expressed as: When the cost difference If the value is greater than or equal to the splitting threshold, it can be confirmed that the identity confirmation of the current candidate pipeline object is incorrect. The candidate identity edge of the current candidate pipeline object is deleted from the current candidate fusion object, and the current candidate pipeline object is split from the current candidate fusion object to update the current candidate fusion object.

[0110] After separating the current candidate pipeline object from the current candidate fusion object, the current candidate pipeline object becomes an independent connection entity. At this point, both this connection entity and the separated candidate fusion object can be considered as updated candidate fusion objects of the current candidate fusion object. Simultaneously, based on the updated candidate fusion objects and the second preset pipeline constraints, the connectivity score between the updated candidate fusion objects is recalculated to determine the updated candidate connection edges between them. Here, the second preset pipeline constraints are the same as the first preset pipeline constraints, and will not be repeated here.

[0111] Furthermore, step 13 above includes:

[0112] Step 133: Update the candidate fusion object as the current candidate fusion object for the next round of iteration update processing, and repeat the process of determining and updating the local connection cost until the iteration number threshold or the preset iteration condition is reached.

[0113] In this embodiment, before the next iteration begins, the candidate identity edges corresponding to multiple candidate network objects in the updated candidate fusion objects are arranged in ascending order, and the two candidate network objects corresponding to the smallest candidate identity edge score are taken as the network objects to be evaluated in the next iteration. The local connection cost of the candidate connection graph when the network object to be evaluated is in the target state is calculated.

[0114] Here, the preset iteration conditions may include that if no invalid identity edges are generated in the current iteration, it is considered that the current update candidate fusion object has reached a stable state and no further update processing is needed, the iteration stops, and all candidate network objects in the final update candidate fusion object have determined the accurate identity of the corresponding entity network object. At the same time, the candidate connection edges between each update candidate fusion object all represent stable topological connection relationships between the update candidate fusion objects.

[0115] In an optional embodiment of the present invention, step 14 above may include:

[0116] Step 141: Based on the standard information record unit of each candidate pipeline object in the updated candidate fusion object, determine the attribute candidate score value and geometric candidate score value of each candidate pipeline object;

[0117] Step 142: Based on the updated candidate connection edges, determine the connection candidate score values ​​for the updated fusion object;

[0118] Step 143: Based on the attribute candidate score, geometric candidate score, and connection candidate score, sort the multiple candidate pipeline objects of the updated candidate fusion object, and determine the primary attribute features, alternative attribute features, primary geometric features, alternative geometric features, connection relationships of physical pipeline objects, and connection relationships to be reviewed for the target information record unit according to the sorting order.

[0119] In this embodiment, based on the data information in the standard information recording unit of each candidate pipeline object in the updated candidate fusion object, the attribute candidate score value and geometric candidate score value of each candidate pipeline object are determined sequentially.

[0120] Preferably, the candidate attribute score can be calculated using the following formula:

[0121]

[0122] in, Q represents the candidate attribute score; Q represents the source data quality value of the attribute features of the current candidate pipeline object (weighted based on the surveying accuracy, acquisition method, result level and data completeness of the source dataset). This represents the time decay value of the attribute feature of the candidate pipeline object (the change of the attribute feature over time, calculated based on the time difference between the acquisition time, completion time, inspection time or version update time of the attribute feature and the current time, the initial value of the attribute feature, and the preset decay coefficient). This indicates the standard semantic confidence level corresponding to the candidate pipeline object; This represents the independent source support of the attribute feature of the candidate pipeline object (the number of source datasets that support this attribute feature). This indicates the historical support of the attribute features of the candidate pipeline object; This represents the conflict penalty value of the attribute feature of the candidate pipeline object (the conflict penalty value is 1 when the attribute feature is inconsistent with the preset value range or conflicts with the preset rule attribute; otherwise, it is 0). These represent the weights of the corresponding parameters, which can be trained or statistically calibrated based on historical fixed attribute samples. When historical samples are insufficient, initial weights are preset according to the importance of source data quality, time decay, semantic confidence, independent source support, historical support, and conflict penalty, and then normalized.

[0123] Preferably, the geometric candidate score can be calculated using the following formula:

[0124]

[0125] in, Q represents the geometric candidate score; Q represents the source data quality value of the geometric features of the candidate pipeline object (weighted based on the surveying accuracy, coordinate datum, result level, acquisition method and data completeness of the source dataset). This represents the normalized value of the registration error of the geometric features of the candidate pipeline object (which can be obtained by normalizing the error between the corresponding data and the corresponding original data after unifying the geometric features of the candidate pipeline object in the GIS pipeline layer, CAD as-built drawings, detection and inspection results, and BIM component model to the target coordinate system). The area or bandwidth normalization value of the geometric uncertainty domain representing the geometric characteristics of the candidate pipeline object; This indicates the topological consistency of the candidate network object after its geometric features are added to the candidate connection graph (whether the geometric features maintain topological continuity, reasonable node degree, and reasonable endpoint connection after being added to the candidate connection graph; if none of the three conditions are met, the topological consistency is 0; if one of the three conditions is met, the topological consistency is 1 / 3; if two conditions are met, the consistency is 2 / 3; and if all three conditions are met, the consistency is 1). This represents the independent source support of the geometric features of the candidate pipeline object (the number of source datasets that support this geometric feature). These represent the weights of the corresponding parameters, which can be obtained through historical sample training, statistical calibration, or expert-preset normalization. Based on the updated candidate connection edges, connection candidate scores can be determined between multiple updated candidate fusion objects. Preferably, the connection candidate scores can be calculated using the following formula:

[0126]

[0127] in, Indicates the connection of candidate score values; The same parameter as the one mentioned in the connectivity score has the same meaning; This indicates the historical support for the candidate connection edge to be updated (which can be obtained by weighting the support for the connection relationship from historical confirmed connection edges, historical maintenance work orders, historical merged versions, or manual review results). These represent the weights of the corresponding parameters, which can be obtained through training based on confirmed connection samples and erroneous connection samples.

[0128] After confirming the attribute candidate scores, geometric candidate scores, and connection candidate scores of all candidate network objects in the updated candidate fusion objects, and for each updated candidate fusion object, the attribute candidate scores of all candidate network objects in each updated candidate fusion object are arranged in descending order. When the highest attribute candidate score is greater than or equal to the attribute setting threshold, and the difference between the highest and second highest attribute candidate scores is greater than or equal to the first difference threshold, the attribute feature corresponding to the highest attribute candidate score is determined as the primary attribute feature, and the attribute feature corresponding to the second highest attribute candidate score is determined as the candidate attribute feature. At the same time, the source dataset identifier corresponding to the primary attribute feature is determined as the primary attribute source identifier, and the source dataset identifier corresponding to the candidate attribute feature is determined as the candidate attribute source identifier.

[0129] For each updated candidate fusion object, the geometric candidate scores of all candidate network objects are arranged in descending order. When the geometric candidate score with the highest value is greater than or equal to the geometric finalization threshold, and the difference between the geometric candidate score with the highest value and the second highest value is greater than or equal to the second difference threshold, the geometric feature corresponding to the geometric candidate score with the highest value is determined as the primary geometric feature, and the geometric feature corresponding to the second highest value is determined as the candidate geometric feature. At the same time, the source dataset identifier corresponding to the primary geometric feature is determined as the primary geometric source identifier, and the source dataset identifier corresponding to the candidate geometric feature is determined as the candidate geometric source identifier.

[0130] For the connection candidate score value of the update candidate connection edge between each update candidate fusion object, if the connection candidate score value of the update candidate connection edge between any two update candidate fusion objects is greater than or equal to the connection determination threshold, then the update candidate connection edge between the two update candidate fusion objects is determined as the connection relationship between the corresponding two entity network objects; otherwise, the update candidate connection edge is determined as the connection relationship to be reviewed.

[0131] like Figure 5 As shown, embodiments of the present invention also provide a multi-source heterogeneous underground pipeline network data processing system 30, comprising:

[0132] The acquisition module is used to acquire multiple source datasets of underground pipe networks;

[0133] The processing module aligns and connects data information from multiple source datasets based on a preset underground pipeline semantic model and preset pipeline constraints, determining candidate connection graphs for pipeline objects within the underground pipeline network. Each vertex in the candidate connection graph represents a candidate fusion object, and each candidate fusion object contains multiple candidate pipeline objects representing the same entity. Candidate connection edges in the candidate connection graph represent candidate connection relationships between two candidate fusion objects. Based on the local connection cost of any candidate pipeline object in the candidate fusion object under the target state, the candidate fusion object and candidate connection edges are iteratively updated to obtain updated candidate fusion objects and updated candidate connection edges. Based on the candidate score values ​​of the candidate pipeline objects and updated candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for each updated candidate fusion object corresponding to an entity pipeline object. The target information recording unit contains the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationships of entity pipeline objects, and connection relationships to be reviewed for the entity pipeline object corresponding to the currently updated candidate fusion object.

[0134] It should be noted that this system is a system corresponding to the above-mentioned method for processing multi-source heterogeneous underground pipeline network data. All implementation methods in the above-mentioned method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0135] like Figure 6 As shown, embodiments of the present invention also provide an electronic device 50, comprising: a memory 51 for storing one or more computer programs; and one or more processors 52 for executing the one or more computer programs. When the computer programs are run by the processors, they perform the processing method for multi-source heterogeneous underground pipeline network data as described in the above embodiments. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects. The electronic device 50 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 50 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown in this invention, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0136] like Figure 7As shown, electronic device 50 is a computing device or computer system, which may include CPU 501 (computing unit), which can perform various appropriate actions and processes according to a computer program stored in ROM 502 (read-only memory) or a computer program loaded from storage unit 508 into random access RAM 503 (memory). RAM 503 may also store various programs and data required for the operation of electronic device 50. CPU 501, ROM 502, and RAM 503 are interconnected via bus 504. I / O interface 505 (input / output interface) is also connected to bus 504.

[0137] Multiple components in electronic device 50 are connected to I / O interface 505, including: input unit 506, such as keyboard, mouse, etc.; output unit 507, such as various types of displays, speakers, etc.; storage unit 508, such as disk, optical disk, etc.; and communication unit 509, such as network card, modem, wireless transceiver, etc. Communication unit 509 allows electronic device 50 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0138] CPU 501 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of CPU 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. CPU 501 performs the various methods and processes described above. For example, in some embodiments, the method for processing multi-source heterogeneous underground pipe network data can be implemented as a computer software program, which is tangibly contained in a computer-readable storage medium, such as storage unit 508. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 50 via ROM 502 and / or communication unit 509. When the computer program is loaded into RAM 503 and executed by CPU 501, one or more steps of the method for processing multi-source heterogeneous underground pipe network data described above can be performed. Alternatively, in other embodiments, CPU 501 may be configured by any other suitable means (e.g., by means of firmware) to perform a method for processing multi-source heterogeneous underground pipeline network data.

[0139] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the methods described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0140] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0141] The units described as separate components may or may not be physically separate. The components shown as units 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 units can be selected to achieve the purpose of this embodiment according to actual needs.

[0142] In addition, the functional units 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.

[0143] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0144] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for processing multi-source heterogeneous underground pipe network data, characterized in that, include: Obtain multiple source datasets of underground pipe networks; Based on a preset underground pipeline semantic model and preset pipeline constraints, the data information of multiple source datasets is aligned and connected to determine the candidate connection graph of pipeline objects in the underground pipeline network. Each vertex in the candidate connection graph represents a candidate fusion object, and the candidate fusion object contains multiple candidate network objects representing the same entity network object; the candidate connection edges in the candidate connection graph represent the candidate connection relationship between two candidate fusion objects. Based on the local connection cost of the candidate connection graph of any candidate pipeline object in the candidate fusion object under the target state, the candidate fusion object and the candidate connection edge are iteratively updated to obtain the updated candidate fusion object and the updated candidate connection edge. Based on the candidate score values ​​of the candidate pipeline objects and the candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the corresponding entity pipeline object of each updated candidate fusion object; The target information recording unit includes the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationship of the physical pipeline object corresponding to the currently updated candidate fusion object, and connection relationship to be reviewed.

2. The method for processing multi-source heterogeneous underground pipeline network data according to claim 1, characterized in that, Based on a preset underground pipeline semantic model and preset pipeline constraints, the data information from multiple source datasets is aligned and connected to determine candidate connection graphs for pipeline objects within the underground pipeline network, including: Based on the preset underground pipeline semantic model, the multiple source datasets are standardized to determine the standard information recording unit for each pipeline object in the underground pipeline network. The standard information recording unit includes the standard semantics of the current pipeline object, the standard semantic confidence, the data information representing the current pipeline object in each source dataset, and the source dataset identifier of the data information. Based on the standard information recording unit of each pipeline object and the first preset pipeline constraint conditions, the candidate connection diagram of the pipeline objects in the underground pipeline network is determined.

3. The method for processing multi-source heterogeneous underground pipeline network data according to claim 2, characterized in that, Based on the preset underground pipeline semantic model, multiple source datasets are standardized to determine the standard information recording unit for each pipeline object in the underground pipeline network, including: Based on multiple source datasets, an initial information recording unit is determined for each pipeline object in the underground pipeline network; the pipeline object includes underground pipe segments and pipeline facilities, and the initial information recording unit includes the original semantic record, data information, and source dataset identifier of the data information representing the pipeline object in each source dataset. Based on the preset underground pipeline semantic model, the original semantic records corresponding to each source dataset in the initial information recording unit of each pipeline object are mapped to obtain the standard semantics and standard semantic confidence of each pipeline object, and a standard information recording unit for each pipeline object is generated; the standard semantic confidence represents the accuracy of mapping the original semantic records to standard semantics.

4. The method for processing multi-source heterogeneous underground pipeline network data according to claim 2, characterized in that, Based on the standard information recording unit of each pipeline object and the first preset pipeline constraint conditions, candidate connection diagrams of pipeline objects in the underground pipeline network are determined, including: Based on the standard information recording unit, the network objects contained in the multiple source datasets are divided to obtain multiple candidate identity graphs. The candidate identity graphs contain multiple candidate network objects representing the same entity, and each candidate network object serves as a vertex of the candidate identity graph. The connecting edges in the candidate identity graphs represent the candidate identity edges of the candidate network objects. The candidate identity graph is used as a candidate fusion object, and based on the candidate fusion object and the first preset pipeline constraint, the candidate connection edge between the two candidate fusion objects is determined. Based on the candidate fusion objects and the candidate connection edges, the candidate connection graph is constructed and obtained.

5. The method for processing multi-source heterogeneous underground pipeline network data according to claim 1, characterized in that, Based on the local connectivity cost of the candidate connection graph of any candidate network object in the candidate fusion objects under the target state, the candidate fusion objects and the candidate connection edges are iteratively updated to obtain updated candidate fusion objects and updated candidate connection edges, including: Determine the first local connection cost of the current candidate connection graph of the current candidate pipeline object in the split state and the second local connection cost of the current candidate connection graph of the current candidate pipeline object in the fusion state; Based on the first local connection cost and the second local connection cost of the current candidate connection graph, the current candidate fusion object and the candidate connection edges in the current candidate connection graph are updated to obtain the updated candidate fusion object of the current candidate fusion object and the updated candidate connection edges of the candidate connection edges in the current candidate connection graph. The updated candidate fusion object is used as the current candidate fusion object for the next round of iterative update processing, and the determination and update process of local connection cost is repeated until the iteration number threshold or preset iteration condition is reached.

6. The method for processing multi-source heterogeneous underground pipeline network data according to claim 5, characterized in that, Determining the first local connectivity cost of the current candidate network object in the split state and the second local connectivity cost of the current candidate network object in the fusion state includes: The cost of local connectivity is determined using the following formula: Where i=1,2 represents the target state, i=1 indicates that the current candidate pipeline object in the current candidate fusion object is in the split state, and i=2 indicates that the current candidate pipeline object in the current candidate fusion object is in the fusion state; Let the cost of local connectivity be denoted as , where the cost of the first local connectivity is denoted as . The cost of the second local connection is expressed as ; This indicates the conflict value between the two ends of the current candidate pipeline object; This indicates the diameter continuity deviation value between the two ends of the current candidate pipeline object; This indicates the elevation interval conflict value between the two ends of the current candidate pipeline object; This indicates the flow direction conflict value at the connection end of the current candidate pipeline object; This indicates an outlier value in the node degree of the current candidate connectivity graph under the target state; This indicates the discontinuity of the node neighborhood in the current candidate connectivity graph under the target state; These represent the weights of the corresponding parameters.

7. The method for processing multi-source heterogeneous underground pipeline network data according to claim 5, characterized in that, Based on the first local connection cost and the second local connection cost of the current candidate connection graph, the current candidate fusion object and the candidate connection edges in the current candidate connection graph are updated to obtain the updated candidate fusion object and the updated candidate connection edges in the current candidate connection graph, including: Determine the cost difference between the first local connection cost and the second local connection cost; in response to the cost difference being greater than or equal to a splitting threshold, determine the candidate identity edge of the current candidate pipeline object as an invalid identity edge. Based on the invalid identity edge, the corresponding current candidate pipeline object in the current candidate fusion object is split to obtain the updated fusion object; Based on the updated fusion object and the second preset pipeline constraint, the candidate connection edge for updating between the two updated fusion objects is determined.

8. The method for processing multi-source heterogeneous underground pipeline network data according to claim 1, characterized in that, Based on the candidate score values ​​of the candidate network objects and the candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the corresponding entity network object of each updated candidate fusion object, including: Based on the standard information recording unit of each candidate pipeline object in the updated candidate fusion object, determine the attribute candidate score value and geometric candidate score value of each candidate pipeline object; Based on the updated candidate connection edges, determine the connection candidate score value of the updated fusion object; Based on the attribute candidate score, the geometric candidate score, and the connection candidate score, the multiple candidate pipeline objects of the updated candidate fusion object are sorted, and the primary attribute features, alternative attribute features, primary geometric features, alternative geometric features, connection relationships of physical pipeline objects, and connection relationships to be reviewed are determined according to the sorting order to generate the target information recording unit.

9. A system for processing multi-source heterogeneous underground pipeline network data, characterized in that, include: The acquisition module is used to acquire multiple source datasets of underground pipe networks; The processing module is used to align and connect the centralized data information of multiple source datasets based on a preset underground pipeline semantic model and preset pipeline constraints, and to determine the candidate connection graph of pipeline objects in the underground pipeline network. Each vertex in the candidate connection graph represents a candidate fusion object, and the candidate fusion object contains multiple candidate network objects representing the same entity network object; the candidate connection edges in the candidate connection graph represent the candidate connection relationship between two candidate fusion objects. Based on the local connection cost of the candidate connection graph of any candidate pipeline object in the candidate fusion object under the target state, the candidate fusion object and the candidate connection edge are iteratively updated to obtain the updated candidate fusion object and the updated candidate connection edge. Based on the candidate score values ​​of the candidate pipeline objects and the candidate connection edges in the updated candidate fusion objects, a target information recording unit is determined for the corresponding entity pipeline object of each updated candidate fusion object; The target information recording unit includes the primary attribute features, alternative attribute features, primary attribute source identifier, alternative attribute source identifier, primary geometric features, alternative geometric features, primary geometric source identifier, alternative geometric source identifier, connection relationship of the physical pipeline object corresponding to the currently updated candidate fusion object, and connection relationship to be reviewed.

10. A computing device, characterized in that, include: Memory, used to store one or more programs; One or more processors are configured to execute the one or more programs to implement the method as described in any one of claims 1 to 8.

Images