Excavation volume statistics method and system based on knowledge graph and three-dimensional geological model

Abstract

This invention proposes a method and system for excavation volume statistics based on knowledge graphs and 3D geological models, relating to the fields of geological engineering and intelligent modeling technology. Addressing the shortcomings of existing technologies, such as the lack of semantic knowledge association in 3D geological models, reliance on manual experience and geometric calculations for volume statistics, inability to automatically identify complex geological bodies and correct parameters, high dependence on manual operation, low level of intelligence, and inability to achieve deep coupling and differentiated measurement between geological knowledge and spatial models, this invention proposes a spatial-semantic bidirectional mapping between a geological engineering knowledge graph and a 3D geological model. This enables automatic identification of abnormal geological bodies, intelligent division of excavation areas, and automatic reasoning and selection of correction coefficients, overcoming the limitations of existing technologies that separate models from knowledge and rely on manual experience. Closed-loop iterative updates are achieved through volume deviation verification and reverse tracing of the knowledge graph, significantly improving the statistical accuracy and intelligence level of excavation volume in complex geological scenarios.

Description

Technical Field

[0001] This invention relates to the fields of geological engineering and intelligent modeling technology, and in particular to a method and system for calculating excavation volume based on knowledge graphs and three-dimensional geological models. Background Technology

[0002] In geotechnical engineering construction, including slope excavation for water conservancy and hydropower projects, open-pit mining, tunnel excavation, and urban deep foundation pit engineering, accurate excavation volume statistics are a core technical aspect of project cost accounting, construction progress control, resource optimization, and cost management. The accuracy of these measurements directly impacts the accuracy of project economic benefit assessments and the rationality of construction decisions. Currently, mainstream excavation volume statistics technologies in the industry primarily rely on methods such as 3D laser point cloud modeling, UAV oblique photogrammetry modeling, BIM modeling, and 3D geological entity modeling.

[0003] However, practical engineering applications have shown that the aforementioned existing technologies have several insurmountable technical defects. First, three-dimensional geological models are essentially pure spatial geometric models, capable of only representing the spatial distribution of rock strata interfaces and topographic relief. They cannot associate with or contain deep semantic knowledge such as soil and rock types, geological structural characteristics, construction technical specifications, historical records of similar work conditions, and engineering experience. Unstructured or semi-structured knowledge such as geological survey reports, construction logs, and specification clauses exist in discrete document form, separated from the three-dimensional model, forming multi-source heterogeneous data silos, resulting in a large amount of valuable engineering experience being unable to be reused in a structured manner. Second, traditional volumetric statistics rely excessively on simple spatial geometric calculations and lack intelligent reasoning and dynamic correction mechanisms. When complex geological structures such as fault fracture zones, karst cave development areas, interlayered soft and hard rocks, or weak interlayers exist within the excavation area, existing methods cannot automatically identify these abnormal geological bodies and adjust measurement strategies and correction parameters accordingly. They can only rely on the personal experience of technical personnel to set coefficients, resulting in large human errors, measurement results that vary from person to person, and difficulty in verification and traceability. Third, key aspects such as the rational division of the excavation area, the identification and removal of abnormal geological bodies, the selection of correction coefficients, and the dynamic correction of the model all require manual completion. This makes it difficult to adapt to the real-time or near-real-time measurement needs of the rapidly changing excavation progress during construction, and it is easily affected by subjective factors, failing to achieve standardization and automation. Fourth, it lacks the ability to deeply couple geological knowledge with the three-dimensional spatial model, making it difficult to perform differentiated automatic measurement based on the geological conditions of different areas within the excavation range and the soil and rock types. The measurement results deviate significantly from the actual situation.

[0004] Therefore, to address the aforementioned technical problems, this application provides an intelligent excavation volume statistics method and system based on knowledge graphs and 3D geological models. By standardizing and preprocessing multi-source heterogeneous geological engineering data, a geological engineering knowledge graph specifically for excavation measurement is constructed. This graph systematically precipitates and expresses geological entities, engineering entities, attribute entities, and their complex relationships in a graph structure. Simultaneously, a high-precision 3D geological model is constructed using hybrid grid technology, establishing a spatial-semantic bidirectional mapping relationship between model grid units and knowledge graph entities. The semantic reasoning capability of the knowledge graph automatically identifies abnormal geological bodies within the excavation area and adaptively divides differentiated measurement sub-regions. Combining historical working condition matching and standard constraints, correction coefficients are automatically selected to achieve intelligent correction of the initial geometric volume and automatic summarization of the total volume. Furthermore, deviation verification of actual measurement data feeds back into the knowledge graph for iterative optimization. This effectively solves the problems of model-knowledge separation, low measurement accuracy in complex geological scenarios, high reliance on manual intervention, and inability to achieve differentiated intelligent measurement in existing technologies. Summary of the Invention

[0005] Existing 3D geological models lack semantic knowledge associations, and volumetric volume statistics rely on manual experience and geometric calculations, failing to automatically identify complex geological bodies and correct parameters. Furthermore, they are highly dependent on manual operation, have low levels of intelligence, and cannot achieve deep coupling between geological knowledge and spatial models, nor differentiated measurement. This invention first proposes a method for excavation volume statistics based on knowledge graphs and 3D geological models, specifically including the following steps: Step S1: Collect multi-source heterogeneous geological engineering data of the target excavation area, and perform standardized preprocessing in sequence, including coordinate unification, format normalization, deduplication, semantic annotation and missing value completion, to construct a standardized geological engineering dataset with spatial coordinates and semantic labels. Step S2: Based on the standardized geological engineering dataset, extract geological entities, engineering entities, attribute entities and their relationships, construct the knowledge graph pattern layer using ontology modeling, and construct a geological engineering knowledge graph with semantic reasoning function based on the graph database. Step S3: Construct a three-dimensional geological model using a hybrid modeling method of TIN triangular mesh and hexahedral mesh. Establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the three-dimensional geological model and the geological entities and attribute entities in the geological engineering knowledge graph, so as to realize the bidirectional call of spatial mesh and semantic knowledge. Step S4: Import the construction design excavation boundary to determine the initial excavation space range, automatically identify abnormal geological bodies within the excavation range through semantic reasoning, adaptively divide differentiated excavation measurement sub-regions, distinguish soft soil areas, bedrock areas, fault-affected areas and karst correction areas, and eliminate non-excavation geological bodies. Step S5: Perform three-dimensional excavation Boolean operations and grid volume integration calculations on each excavation measurement sub-area to obtain the initial geometric volume of each sub-area. Automatically select the corresponding correction coefficient by matching historical similar working conditions and engineering specifications through knowledge graph, dynamically correct the initial geometric volume, and accumulate the corrected volumes of each sub-area to obtain the total excavation volume of the target area. Step S6: Verify the deviation between the total excavation volume calculated by the intelligent system and the actual on-site measurement data. If the deviation exceeds the error threshold, trace the source of the deviation through the knowledge graph and feed back the verification results, deviation correction rules and new engineering data to the knowledge graph to update entity attributes and reasoning rules, thereby achieving iterative optimization of the knowledge graph.

[0006] Furthermore, in step S1, the multi-source heterogeneous geological engineering data includes geological exploration borehole data, rock and soil mechanics parameters, geological structure data, three-dimensional topographic point cloud data, construction design excavation boundary data, engineering construction specification data, historical excavation condition data of similar projects, and geological disaster risk data.

[0007] Furthermore, in step S2, the geological entities include rock strata, soil types, faults, karst caves, and weak interlayers; the engineering entities include excavation boundaries, construction procedures, and measurement units; the attribute entities include soil and rock density, compressive strength, loosening coefficient, over-excavation coefficient, and compaction coefficient; and the relationships include geological inclusion relationships, soil and rock coefficient mapping relationships, specification constraint relationships, and historical working condition analogy relationships.

[0008] Furthermore, in step S2, the geological engineering knowledge graph supports forward semantic reasoning, rule matching, and reverse error tracing.

[0009] Furthermore, in step S5, the correction coefficients are automatically selected by matching historical working conditions and engineering specifications using the knowledge graph, without the need for manual setting, and the correction coefficients include the loosening coefficient, over-excavation coefficient, and compaction coefficient.

[0010] According to another aspect of the present invention, an excavation volume statistics system based on knowledge graphs and three-dimensional geological models is also proposed, applied to the aforementioned excavation volume statistics method based on knowledge graphs and three-dimensional geological models, comprising: The data acquisition and preprocessing module is used to collect multi-source heterogeneous geological engineering data of the target excavation area and perform standardized preprocessing such as coordinate unification, format normalization, deduplication, semantic annotation and missing value completion. The knowledge graph construction and reasoning module is used to extract geological entities, engineering entities, attribute entities and their relationships. It uses ontology modeling to construct the knowledge graph pattern layer and builds a geological and engineering knowledge graph with semantic reasoning capabilities based on a graph database. The 3D geological modeling module is used to construct 3D geological models using a hybrid modeling method of TIN triangular mesh and hexahedral mesh, and to establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the 3D geological model and the geological entities and attribute entities in the knowledge graph. The intelligent excavation area division module is used to import the construction design excavation boundary and automatically identify abnormal geological bodies through semantic reasoning of knowledge graph, and adaptively divide differentiated excavation measurement sub-areas. The intelligent excavation volume calculation module is used to perform three-dimensional excavation Boolean operations and grid volume integration calculations on each excavation measurement sub-area. It automatically selects correction coefficients through knowledge graph reasoning to complete volume correction and summarizes the total excavation volume. The verification and iteration module is used to verify the deviation between the calculated volume and the actual on-site measurement data, and to feed the verification results and correction rules back to the knowledge graph to achieve iterative optimization.

[0011] Furthermore, the knowledge graph construction and reasoning module supports forward semantic reasoning, rule matching, and reverse error tracing.

[0012] Furthermore, the 3D geological modeling module constructs a 3D geological model in which each individual grid cell is bound to a geological entity and attribute entity in the knowledge graph, enabling direct access to the corresponding geological attributes and measurement parameters in the knowledge graph through the grid cell.

[0013] Furthermore, the intelligent excavation area division module distinguishes differentiated excavation measurement sub-regions, including soft soil areas, bedrock areas, fault-affected areas, and karst correction areas, through semantic reasoning based on knowledge graphs, and automatically masks invalid measurement ranges.

[0014] Furthermore, when the deviation between the calculated volume and the actual on-site measurement data exceeds the error threshold, the verification iteration module uses the knowledge graph to construct the reasoning module to trace back and locate the cause of the deviation, and feeds back the deviation correction rules and newly added engineering data to the knowledge graph to update the entity attributes and reasoning rules.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: First, it deeply integrates spatial and semantic knowledge, breaking through the limitations of pure geometric measurement. This invention constructs a geological engineering knowledge graph and establishes a spatial-semantic bidirectional mapping with a 3D geological model, enabling each grid cell to be associated with semantic knowledge such as soil and rock types, geological structures, and correction coefficients. This changes the situation in existing technologies where the 3D model only serves as a geometric carrier, realizing deep coupling and knowledge reuse of geological knowledge and spatial models. Secondly, the entire process features intelligent reasoning, eliminating reliance on human experience. The semantic reasoning capability based on knowledge graphs in this invention automatically identifies abnormal geological bodies such as faults and karst caves and adaptively divides the excavation measurement sub-zones. At the same time, it automatically selects correction coefficients to complete the volume correction by matching historical working conditions and engineering specifications. No manual parameter setting is required throughout the entire process, which significantly improves measurement accuracy and statistical efficiency compared to existing technologies. Thirdly, closed-loop iterative updates support dynamic traceability and optimization. This invention verifies the discrepancies between calculated quantities and on-site measured data, and uses a knowledge graph to trace the causes of deviations in reverse. Correction rules and new data are then fed back to the knowledge graph for iterative updates. This overcomes the shortcomings of existing technologies in dynamically optimizing and tracing data, providing traceable technical support for engineering auditing and measurement payment. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 A flowchart of an excavation volume statistics method based on knowledge graphs and 3D geological models; Figure 2 This is a schematic diagram of the excavation volume statistics system based on knowledge graphs and three-dimensional geological models. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0019] The specific embodiments of the present invention will be described below.

[0020] To address the shortcomings of existing technologies, such as the lack of semantic knowledge association in 3D geological models, reliance on manual experience and geometric calculations for volume statistics, and the inability to automatically identify complex geological bodies and correct parameters, this invention proposes a spatial-semantic bidirectional mapping between a geological engineering knowledge graph and a 3D geological model. This enables automatic identification of abnormal geological bodies, intelligent division of excavation areas, and automatic reasoning and selection of correction coefficients, fundamentally overcoming the limitations of existing technologies that separate models from knowledge and rely on manual experience. Furthermore, closed-loop iterative updates are achieved through volume deviation verification and reverse tracing of the knowledge graph, significantly improving the statistical accuracy and intelligence level of excavation volume in complex geological scenarios.

[0021] Example 1 like Figure 1 As shown, this invention proposes a method for excavation volume statistics based on knowledge graphs and 3D geological models. By constructing a spatial-semantic bidirectional mapping between a geological engineering knowledge graph and a 3D geological model, it achieves automatic identification of abnormal geological bodies, intelligent division of excavation areas, automatic inference and selection of correction coefficients, and closed-loop iterative updates. This fundamentally solves the problems of model-knowledge separation, reliance on human experience, and low measurement accuracy in complex geological scenarios in existing technologies, including: Step S1: Collect multi-source heterogeneous geological engineering data of the target excavation area, and perform standardized preprocessing in sequence, including coordinate unification, format normalization, deduplication, semantic annotation, and missing value completion, to construct a standardized geological engineering dataset with spatial coordinates and semantic labels. The multi-source heterogeneous geological engineering data includes geological exploration borehole data, geotechnical mechanics parameters, geological structure data, 3D topographic point cloud data, construction design excavation boundary data, engineering construction specification data, historical excavation condition data of similar projects, and geological hazard risk data.

[0022] Specifically, the collected data undergoes standardized preprocessing to eliminate heterogeneous data format differences, remove duplicate data to eliminate redundant information, assign geological meaning labels to each data point through semantic annotation, and fill in missing values ​​to ensure data integrity. Finally, a standardized geological engineering dataset carrying both spatial coordinate information and semantic labels is constructed.

[0023] Step S2: Based on a standardized geological engineering dataset, geological entities, engineering entities, attribute entities, and their relationships are extracted. An ontology modeling layer is used to construct a knowledge graph schema layer, and a geological engineering knowledge graph with semantic reasoning capabilities is built based on a graph database. Geological entities include rock strata, soil types, faults, karst caves, and weak interlayers; engineering entities include excavation boundaries, construction procedures, and measurement units; attribute entities include soil and rock density, compressive strength, loosening coefficient, over-excavation coefficient, and compaction coefficient; and relationships include geological inclusion relationships, soil-coefficient mapping relationships, specification constraint relationships, and historical working condition analogy relationships. The geological engineering knowledge graph supports forward semantic reasoning, rule matching, and reverse error tracing.

[0024] An ontology modeling approach is used to construct the schema layer of the knowledge graph, clearly defining the types and attributes of various entities and relationships. Based on the Neo4j graph database, the aforementioned entities and relationships are stored in a graph structure, constructing a geological engineering knowledge graph with forward semantic reasoning, rule matching, and reverse error tracing capabilities.

[0025] Step S3: Construct a three-dimensional geological model using a hybrid modeling method of TIN triangular mesh and hexahedral mesh. Establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the three-dimensional geological model and the geological entities and attribute entities in the geological engineering knowledge graph, so as to realize the bidirectional call of spatial mesh and semantic knowledge.

[0026] Specifically, the TIN triangular mesh is used to characterize rock strata interfaces, geological structural boundaries, and topographic relief surfaces, while the hexahedral mesh is used to fill the interior of rock strata and provide a detailed representation of local complex structures. This mapping allows for the direct retrieval of semantic knowledge of the corresponding geological type and measurement parameters from any spatial grid cell; conversely, it also enables the location of the spatial grid region covered by any entity in the knowledge graph, thus achieving bidirectional access between spatial grids and semantic knowledge.

[0027] Step S4: Import the construction design excavation boundary to determine the initial excavation space. Automatically identify anomalous geological bodies within the excavation area through semantic reasoning, adaptively divide the area into differentiated excavation measurement sub-regions, distinguishing between soft soil areas, bedrock areas, fault-affected areas, and karst correction areas, and eliminating non-excavation geological bodies. Specifically, semantic reasoning based on the constructed knowledge graph automatically identifies anomalous geological bodies within the initial excavation area, including non-uniform geological structures such as faults, karst caves, and weak interlayers. Based on the reasoning and identification results, the excavation area is adaptively divided into differentiated excavation measurement sub-regions, distinguishing between areas with different geological conditions such as soft soil areas, bedrock areas, fault-affected areas, and karst correction areas, while automatically eliminating non-excavation geological bodies and masking invalid measurement ranges.

[0028] Step S5: Perform 3D excavation Boolean operations and mesh volume integral calculations on each excavation measurement sub-zone to obtain the initial geometric volume of each sub-zone. Automatically select corresponding correction coefficients by matching historical similar working conditions with engineering specifications using a knowledge graph, dynamically correcting the initial geometric volume. The corrected volumes of each sub-zone are then summed to obtain the total excavation volume of the target area. The correction coefficients are automatically selected by matching historical working conditions with engineering specifications using a knowledge graph, requiring no manual setting. These correction coefficients include a loosening coefficient, an over-excavation coefficient, and a compaction coefficient. For each of the divided excavation measurement sub-zones, 3D excavation Boolean operations and mesh volume integral calculations are performed to obtain the initial geometric volume of each sub-zone. The corresponding correction coefficients, including the loosening coefficient, over-excavation coefficient, and compaction coefficient, are automatically selected by matching historical similar engineering working condition data with engineering construction specification constraints using a knowledge graph. The entire process requires no manual parameter setting. The initial geometric volume of each sub-zone is dynamically corrected using the automatically selected correction coefficients. The corrected volumes of each sub-zone are then summed to obtain the total excavation volume of the target area.

[0029] Step S6: Verify the deviation between the total excavation volume calculated by the intelligent system and the actual on-site measurement data. If the deviation exceeds the error threshold (typically ±10%), the cause of the deviation is located through reverse tracing using the knowledge graph. The verification results, deviation correction rules, and newly added engineering data are fed back to the knowledge graph to update entity attributes and inference rules, thereby achieving iterative optimization of the knowledge graph. Specifically, if the deviation value exceeds the preset error threshold, the specific cause of the deviation is located step-by-step along the inference path using the reverse tracing function of the knowledge graph.

[0030] In this embodiment, spatial and semantic elements are deeply integrated, overcoming the limitations of pure geometric measurement. By constructing a geological engineering knowledge graph and establishing a spatial-semantic bidirectional mapping with the 3D geological model, each grid cell can be associated with semantic knowledge such as soil type, geological structure, and correction coefficients. This changes the situation in existing technologies where the 3D model only serves as a geometric carrier, realizing deep coupling and knowledge reuse of geological knowledge and spatial models. Intelligent reasoning throughout the entire process eliminates reliance on manual experience. Simultaneously, based on the semantic reasoning capabilities of the knowledge graph, it automatically identifies abnormal geological bodies such as faults and karst caves and adaptively divides excavation measurement sub-zones. Furthermore, by matching historical working conditions and engineering specifications, it automatically selects correction coefficients to complete volume correction. No manual parameter setting is required throughout the process, significantly improving measurement accuracy and statistical efficiency compared to existing technologies.

[0031] Example 2 like Figure 2 As shown, this invention also proposes an excavation volume statistics system based on knowledge graphs and three-dimensional geological models, using the excavation volume statistics method based on knowledge graphs and three-dimensional geological models as described in Example 1, including the following: The data acquisition and preprocessing module is used to collect multi-source heterogeneous geological engineering data of the target excavation area and perform standardized preprocessing such as coordinate unification, format normalization, deduplication, semantic annotation and missing value completion. The knowledge graph construction and reasoning module is used to extract geological entities, engineering entities, attribute entities and their relationships. It uses ontology modeling to construct the knowledge graph pattern layer and builds a geological and engineering knowledge graph with semantic reasoning capabilities based on a graph database. The 3D geological modeling module is used to construct 3D geological models using a hybrid modeling method of TIN triangular mesh and hexahedral mesh, and to establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the 3D geological model and the geological entities and attribute entities in the knowledge graph. The intelligent excavation area division module is used to import the construction design excavation boundary and automatically identify abnormal geological bodies through semantic reasoning of knowledge graph, and adaptively divide differentiated excavation measurement sub-areas. The intelligent excavation volume calculation module is used to perform three-dimensional excavation Boolean operations and grid volume integration calculations on each excavation measurement sub-area. It automatically selects correction coefficients through knowledge graph reasoning to complete volume correction and summarizes the total excavation volume. The verification and iteration module is used to verify the deviation between the calculated volume and the actual on-site measurement data, and to feed the verification results and correction rules back to the knowledge graph to achieve iterative optimization.

[0032] The knowledge graph construction and reasoning module supports forward semantic reasoning, rule matching, and reverse error tracing. The 3D geological modeling module constructs a 3D geological model where each grid cell is bound to a geological entity and attribute entity in the knowledge graph, enabling direct access to the corresponding geological attributes and measurement parameters from the knowledge graph through the grid cell. The intelligent excavation area division module uses knowledge graph semantic reasoning to differentiate excavation measurement sub-regions, including soft soil areas, bedrock areas, fault-affected areas, and karst correction areas, and automatically masks invalid measurement ranges. The verification and iteration module, when the deviation between the calculated volume and the actual on-site measurement data exceeds the error threshold, uses the knowledge graph construction and reasoning module to reverse-trace the cause of the deviation and feeds back the deviation correction rules and newly added engineering data to the knowledge graph to update entity attributes and reasoning rules.

[0033] In this embodiment, the system includes interconnected data acquisition and preprocessing modules, a knowledge graph construction and reasoning module, a 3D geological modeling module, an intelligent excavation area division module, an intelligent excavation volume calculation module, and a verification and iteration module. The data acquisition and preprocessing module collects multi-source heterogeneous geological engineering data from the target excavation area, including geological exploration borehole data, geotechnical parameters, geological structure data, 3D topographic point cloud data, construction design excavation boundary data, engineering construction specification data, historical excavation condition data of similar projects, and geological hazard risk data. It performs standardized preprocessing on the collected data, including coordinate unification, format normalization, deduplication, semantic annotation, and missing value completion. The knowledge graph construction and reasoning module uses the Neo4j graph database to extract geological entities, engineering entities, attribute entities, and their relationships. It uses ontology modeling to construct the knowledge graph schema layer and builds a geological engineering knowledge graph with forward semantic reasoning, rule matching, and reverse error tracing functions. The 3D geological modeling module constructs a 3D geological model using a hybrid modeling approach combining TIN triangular meshes and hexahedral meshes. It binds each individual mesh cell of the 3D geological model to geological entities and attribute entities in the knowledge graph, enabling a spatial-semantic bidirectional mapping between the corresponding geological attributes and measurement parameters in the knowledge graph through direct access to the mesh cells. The intelligent excavation area division module imports the excavation boundaries from the construction design. Through semantic reasoning from the knowledge graph, it automatically identifies anomalous geological bodies and adaptively divides differentiated excavation measurement sub-zones, including soft soil zones, bedrock zones, fault-affected zones, and karst correction zones, while automatically masking invalid measurement ranges. The intelligent excavation volume calculation module performs 3D excavation simulation, 3D excavation Boolean operations, and mesh volume integration calculations for each excavation measurement sub-zone. It automatically selects correction coefficients through knowledge graph reasoning to correct the volume and summarizes the total excavation volume. The verification and iteration module is used to verify the deviation between the calculated volume and the actual on-site measurement data. When the deviation exceeds the error threshold, the reasoning module is built through the knowledge graph to trace back to locate the cause of the deviation, and feeds back the deviation correction rules and new engineering data to the knowledge graph to update the entity attributes and reasoning rules, so as to realize the closed-loop iterative update of the system.

[0034] This embodiment verifies the discrepancy between the calculated volume and the actual measured data on site, and uses a knowledge graph to trace the source of the discrepancy. The correction rules and new data are fed back to the knowledge graph to achieve iterative updates. This overcomes the shortcomings of existing technologies that cannot be dynamically optimized and traced, and provides traceable technical support for engineering auditing and measurement payment.

[0035] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present invention.

Claims

1. A method for calculating excavation volume based on knowledge graphs and three-dimensional geological models, characterized in that, Includes the following steps: Step S1: Collect multi-source heterogeneous geological engineering data of the target excavation area, and perform standardized preprocessing on the data after coordinate unification, format normalization, deduplication, semantic annotation and missing value completion in sequence to construct a standardized geological engineering dataset with spatial coordinates and semantic labels. Step S2: Based on the standardized geological engineering dataset, extract geological entities, engineering entities, attribute entities and their relationships, construct a knowledge graph pattern layer using ontology modeling, and construct a geological engineering knowledge graph with semantic reasoning function based on the graph database. Step S3: Construct a three-dimensional geological model using a hybrid modeling method of TIN triangular mesh and hexahedral mesh. Establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the three-dimensional geological model and the geological entities and attribute entities in the geological engineering knowledge graph, so as to realize the bidirectional call of spatial mesh and semantic knowledge. Step S4: Import the construction design excavation boundary to determine the initial excavation space range. Automatically identify abnormal geological bodies within the excavation range through semantic reasoning, adaptively divide differentiated excavation measurement sub-regions, distinguish soft soil areas, bedrock areas, fault-affected areas and karst correction areas, and eliminate non-excavation geological bodies. Step S5: Perform three-dimensional excavation Boolean operations and grid volume integration calculations on each excavation measurement sub-area to obtain the initial geometric volume of each sub-area. Automatically select the corresponding correction coefficient by matching historical similar working conditions and engineering specifications through knowledge graph, dynamically correct the initial geometric volume, and accumulate the corrected volumes of each sub-area to obtain the total excavation volume of the target area. Step S6: Verify the deviation between the total excavation volume calculated by the intelligent system and the actual on-site measurement data. If the deviation exceeds the error threshold, trace the source of the deviation through the knowledge graph and feed back the verification results, deviation correction rules and new engineering data to the knowledge graph to update entity attributes and reasoning rules, thereby achieving iterative optimization of the knowledge graph. 2.The knowledge graph and three-dimensional geological model based excavation volume statistics method of claim 1, wherein, In step S1, the multi-source heterogeneous geological engineering data includes geological exploration borehole data, rock and soil mechanics parameters, geological structure data, three-dimensional topographic point cloud data, construction design excavation boundary data, engineering construction specification data, historical excavation condition data of similar projects, and geological disaster risk data. 3.The knowledge graph and three-dimensional geological model based excavation volume statistics method of claim 2, characterized in that, In step S2, the geological entities include rock strata, soil types, faults, karst caves, and weak interlayers; the engineering entities include excavation boundaries, construction procedures, and measurement units; the attribute entities include soil density, compressive strength, loosening coefficient, over-excavation coefficient, and compaction coefficient; and the correlation relationships include geological inclusion relationships, soil-coefficient mapping relationships, specification constraint relationships, and historical working condition analogy relationships.

4. The knowledge graph and three-dimensional geological model-based excavation volume statistics method according to claim 3, characterized in that, In step S2, the geological engineering knowledge graph supports forward semantic reasoning, rule matching, and reverse error tracing.

5. The method for calculating excavation volume based on knowledge graph and three-dimensional geological model according to claim 1, characterized in that, In step S5, the correction coefficient is automatically selected by matching historical working conditions and engineering specifications using a knowledge graph, without the need for manual setting, and the correction coefficient includes a loosening coefficient, an over-excavation coefficient, and a compaction coefficient.

6. A knowledge graph and three-dimensional geological model-based excavation volume statistics system, applicable to the excavation volume statistics method based on knowledge graph and three-dimensional geological model as described in any one of claims 1 to 5, characterized in that, include: The data acquisition and preprocessing module is used to collect multi-source heterogeneous geological engineering data of the target excavation area and perform standardized preprocessing such as coordinate unification, format normalization, deduplication, semantic annotation and missing value completion. The knowledge graph construction and reasoning module is used to extract geological entities, engineering entities, attribute entities and their relationships. It uses ontology modeling to construct the knowledge graph pattern layer and builds a geological and engineering knowledge graph with semantic reasoning capabilities based on a graph database. The 3D geological modeling module is used to construct 3D geological models using a hybrid modeling method of TIN triangular mesh and hexahedral mesh, and to establish a unique bidirectional mapping relationship between the individual volume mesh or triangular mesh unit of the 3D geological model and the geological entities and attribute entities in the knowledge graph. The intelligent excavation area division module is used to import the construction design excavation boundary and automatically identify abnormal geological bodies through semantic reasoning of knowledge graph, and adaptively divide differentiated excavation measurement sub-areas. The intelligent excavation volume calculation module is used to perform three-dimensional excavation Boolean operations and grid volume integration calculations on each excavation measurement sub-area. It automatically selects correction coefficients through knowledge graph reasoning to complete volume correction and summarizes the total excavation volume. The verification and iteration module is used to verify the deviation between the calculated volume and the actual on-site measurement data, and to feed the verification results and correction rules back to the knowledge graph to achieve iterative optimization.

7. The knowledge graph and three-dimensional geological model-based excavation volume statistics system according to claim 6, characterized in that, The knowledge graph construction and reasoning module supports forward semantic reasoning, rule matching, and reverse error tracing. 8.The knowledge graph and three-dimensional geological model based excavation volume statistics system of claim 6, The three-dimensional geological model constructed by the three-dimensional geological modeling module has individual grid cells that are bound one-to-one with geological entities and attribute entities in the knowledge graph, enabling direct access to the geological attributes and measurement parameters corresponding to the knowledge graph through the grid cells. 9.The knowledge graph and three-dimensional geological model based excavation volume statistics system according to claim 6, characterized in that, The intelligent excavation area division module uses knowledge graph semantic reasoning to distinguish differentiated excavation measurement sub-regions, including soft soil areas, bedrock areas, fault-affected areas, and karst correction areas, and automatically shields invalid measurement ranges. 10.The knowledge graph and three-dimensional geological model based excavation volume statistics system according to claim 6, characterized in that, When the deviation between the calculated volume and the actual on-site measurement data exceeds the error threshold, the verification iteration module uses the knowledge graph to construct the reasoning module to trace back and locate the cause of the deviation, and feeds back the deviation correction rules and newly added engineering data to the knowledge graph to update the entity attributes and reasoning rules.

Images