Water conservancy design decision support system based on knowledge graph

Abstract

The application relates to the technical field of engineering design management, in particular to a water conservancy design decision support system based on a knowledge graph, which comprises a parameter constraint construction module, a design standard fusion module, a task semantic attribution module, a path screening and closing module and a structure combination response module.In the application, the logical linkage organization among structure parameters is realized by constructing a parameter control path and combining multiple relationship recognitions among structure nodes, the constraint recognition capability for the structure control path is enhanced by using a cross relationship judgment mechanism among standard clauses, the structure path screening is provided with a semantic guiding feature by combining the index relationship of the task semantic label, the structure matching precision of the path screening is optimized by using a continuity judgment method of the closed path, the collaborative adaptability among multiple structures is improved by using a response construction logic of the structure combination relationship, the linkage path with the continuity and the logical attachment features is established, and the logical integrity and the response coordination of the water conservancy design are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of engineering design management, and particularly relates to a water conservancy design decision support system based on a knowledge graph. BACKGROUND

[0002] The technical field of engineering design management includes systematic organization and management of various types of information, resources, processes and tasks involved in the whole process of engineering projects from planning and design to implementation, including design scheme formulation, project task allocation, design resource scheduling, design process monitoring, version management, design collaboration and the like, aiming to improve engineering design efficiency and quality through scientific management means under the premise of meeting functional and specification requirements. In actual engineering, engineering design management technology combines information systems and modeling tools to build a decision support model oriented to the design process, assisting designers in multi-source data analysis, design logic analysis and scheme formulation. It also involves structured modeling of design knowledge, standardized management of design standards and establishment of cross-professional collaboration mechanisms. Through the integration of information modeling, knowledge organization and intelligent reasoning methods, informationization and intelligentization support for engineering design activities are realized. The water conservancy design decision support system based on a knowledge graph refers to the construction of a knowledge model system in the field of water conservancy engineering design using knowledge graph technology, and the provision of auxiliary decision-making capabilities required in the design phase based on the model. For technical matters such as design specifications, structure types, hydrological parameters, geological conditions and construction schemes involved in water conservancy design, entity extraction and relationship modeling are performed based on water conservancy engineering standards and historical project data to construct a water conservancy design knowledge graph. Subsequently, design task matching, parameter recommendation and structure combination analysis are performed through the semantic relationship network in the graph, and the optimal design path is inferred based on design working condition requirements. Through entity recognition, relationship extraction, graph construction and reasoning engine, the organization, reasoning and application support of water conservancy design knowledge are realized, and knowledge association and judgment basis are provided in the design management process.

[0003] Traditional water conservancy design decision support technology usually relies on static knowledge modeling methods when dealing with water conservancy design processes, and fails to effectively quantify the linkage constraints between structure nodes and parameters. In the face of the uncertainty of cross-influence and semantic control path between multiple nodes in structure design scenarios, it is difficult to determine the control direction and influence path between parameters. In the standard clause adaptation process, there is a problem of insufficient control relationship coverage. In the task semantic control path generation, the mapping structure between the main control nodes and functional parameters is not introduced, resulting in low path matching accuracy and lack of associated judgment mechanism based on actual working conditions in structure combination analysis. In the construction of the relationship between dam inclination and diversion section, there is a lack of proportion quantization and path continuity verification mechanism, affecting the establishment integrity of structure response path. SUMMARY

[0004] The application aims at solving the defects in the prior art and proposes a water conservancy design decision support system based on a knowledge graph.

[0005] To achieve the above-mentioned purpose, the application adopts the following technical scheme: the water conservancy design decision support system based on a knowledge graph comprises:

[0006] The parameter constraint construction module calls an engineering survey data interface, analyzes engineering properties of various survey data and action paths in design, calculates linkage relationships between rock strength and lining thickness and establishes a parameter control path, calibrates constraint transmission chains between structure nodes, and obtains a parameter logical attachment structure;

[0007] The design standard fusion module uses the parameter logical attachment structure, analyzes control boundaries between structure nodes and design parameters, retrieves constraint clauses of structure parameters in standard texts, calculates control combinations between controlled parameters and establishes a connection path, identifies structure nodes with common control paths, and obtains an index logical fusion density;

[0008] The task semantic attribution module analyzes functional distribution of structure nodes in task descriptions based on the index logical fusion density, identifies dominant parameter control paths and matches task semantic keywords, establishes an index relationship between semantic labels and master nodes, and obtains master semantic identification labels;

[0009] The path screening closure module analyzes path parameter sets associated with master nodes in the graph based on the master semantic identification labels, calculates continuous connection relationships between parameters and screens an effective path set, and obtains path closure structure records;

[0010] The structure combination response module analyzes multi-structure combination relationships in continuous path segments based on the path closure structure records, extracts structure nodes including dam bodies, diversion channels and water discharge facilities, calculates geometric correspondence and control proportions between structures, establishes response connection paths and records path information, and obtains combination response coordination information.

[0011] As a further scheme of the application, the parameter logical attachment structure comprises water head parameter control boundaries, seepage strength linkage relationships and structure node attachment paths, the index logical fusion density is specifically structure parameter control number distribution, standard clause control level and clause boundary intersection identification, the master semantic identification label comprises function control labels, semantic master path identification and task entry mapping sequence, the path closure structure record is specifically structure path connection information, boundary alignment matching order and path number screening list, and the combination response coordination information comprises structure node response sequence, combination control proportion relationship and connection path index information.

[0012] As a further scheme of the application, the parameter constraint construction module comprises:

[0013] The survey parameter identification submodule calls the engineering survey data interface, analyzes the engineering properties and role paths in the design of various survey data, including water head height, permeability coefficient and rock strength, judges the coupling trend of the change interval of the permeability coefficient at the stratum boundary and the change rate of the rock strength, establishes an engineering property matching table of hydrological permeability characteristics and rock mass structure response, and obtains a survey parameter response relationship reference value;

[0014] The control relationship derivation submodule analyzes the stability of the response of the change of rock strength to the thickness of the lining according to the survey parameter response relationship reference value, calculates the boundary coupling state among rock strength, water head height and permeability coefficient, calculates the linkage structure constraint strength value, judges the response conduction direction between linkage structures, and generates a linkage parameter action coefficient;

[0015] The path dependence construction submodule analyzes the logical dependence order of the control path and establishes the directional connection between nodes based on the linkage parameter action coefficient, records the parameter source information associated with the control path, identifies and calibrates the constraint conduction chain between structure nodes, and obtains a parameter logical dependence structure.

[0016] As a further scheme of the present application, the design standard fusion module comprises:

[0017] The clause attribution determination submodule obtains the parameter logical dependence structure, analyzes the control boundary between the structure nodes and the design parameters, retrieves the constraint clauses of the structure parameters in the standard text, including the thickness of the protective layer, the width of the foundation and the cross-section size, judges the control attribution relationship of the standard items corresponding to the structure nodes, identifies the parameter corresponding mode between the structure nodes and the design clauses, establishes the attribution distribution between the structure nodes and the standard clauses, and obtains a standard attribution distribution value;

[0018] The parameter connection generation submodule calculates the control direction combination between the controlled parameters and establishes the parameter connection path based on the standard attribution distribution value, combines the parameter connection paths between multiple structure nodes, calculates the connection of each group of parameters, and generates a parameter connection combination coefficient;

[0019] The index cross calculation submodule calls the parameter connection combination coefficient, compares the cross situation of the parameter control content in each standard clause, identifies the structure nodes with common control paths and records the associated clause numbers, analyzes the multidirectional control relationship of the design standards in the graph structure, and obtains an index logical fusion density.

[0020] As a further scheme of the present application, the process of establishing the ownership distribution between the structure node and the standard clause is specifically: after identifying the control ownership relationship of the structure node corresponding to the standard item, a parameter set related to the protection layer thickness, the foundation width and the cross section size is extracted for each structure node, the association ratio of the node parameter and the clause constraint content is calculated according to the occurrence frequency and the reference position of each parameter in the standard clause text, the ownership determination threshold is set according to the distribution range of the association ratio, and the ownership determination threshold is obtained by calculating the difference between the average constraint strength of the node parameter and the boundary constraint value of the same parameter item in the standard clause;

[0021] In the ownership calculation process, whether the association ratio of each node parameter exceeds the ownership determination threshold is judged, if yes, the target node and the corresponding standard clause are connected in ownership, and the serial number information of the target connection in the ownership distribution is recorded, the ownership distribution between each node and the standard clause is generated according to the number of ownership connections and the association ratio weight between the nodes and the standard clause, and the standard ownership distribution value is obtained.

[0022] As a further scheme of the present application, the task semantic ownership module comprises:

[0023] The function distribution extraction sub-module analyzes the function content of the structure node in the task description based on the index logical fusion density, retrieves the function parameter content associated with the node, including the storage capacity, the flow rate regulation and the outflow rhythm, analyzes the distribution of each parameter in the structure node, identifies the function ownership partition of the node in the task scene, and obtains the node function distribution coefficient;

[0024] The parameter influence determination sub-module calculates the influence direction of each structure parameter in the control path according to the node function distribution coefficient, compares the function control difference of multiple structure nodes in the task path, identifies the dominant parameter and constructs the master parameter path, and obtains the dominant parameter influence path value;

[0025] The semantic label mapping sub-module calls the dominant parameter influence path value, identifies the task semantic keywords corresponding to the dominant parameter, establishes the index relationship between the semantic label and the master node, constructs the atlas task entry mapping and records the semantic label binding information, and obtains the master semantic recognition label.

[0026] As a further scheme of the present application, the path screening and closing module comprises:

[0027] The path parameter extraction sub-module analyzes the path parameter set associated with the master node in the atlas based on the master semantic recognition label, retrieves the structure node related to the channel slope, the storage section and the cross section width, numbers each path segment and records the parameter index, and obtains the path structure parameter sequence;

[0028] The boundary connection judgment sub-module judges the continuous connection relationship between the structure segment parameters according to the path structure parameter sequence, judges the order and connectivity of the boundary intersection in the parameter transmission path, calculates the path continuous matching degree value, compares the continuous matching degree value with the structure segment connection standard value, and obtains the path structure alignment index;

[0029] The effective path screening sub-module identifies the nodes of path breakage and boundary mismatch according to the path structure alignment index, records the path number, eliminates the path segment missing the synchronous control relationship, screens the effective path set, establishes the continuity verification chain, and obtains the path closed structure record.

[0030] As a further scheme of the application, the structure combination response module comprises:

[0031] The structure relationship extraction sub-module analyzes the multi-structure combination relationship in the continuous path segment based on the path closed structure record, extracts the structure nodes including the dam body, the diversion channel and the discharge facility, counts the combination distribution of the structure nodes, identifies the structure combination relationship, and obtains the structure node combination distribution table;

[0032] The parameter linkage judgment sub-module calculates the corresponding proportion between the dam body inclination angle and the diversion channel section shape according to the structure node combination distribution table, analyzes the influence direction of the inclination angle change on the diversion section geometric characteristics, judges the control parameter of the channel section change on the discharge facility flow capacity, and obtains the parameter linkage response value;

[0033] The response path generation sub-module calls the parameter linkage response value, establishes the response connection path between the diversion nodes and the discharge nodes, identifies the parameter transmission order between the nodes in the path, analyzes the stability of the parameter control transmission between the nodes, identifies the structure combination forming the linkage response and records the path information, and obtains the combination response coordination information.

[0034] Compared with the prior art, the application has the advantages and positive effects that:

[0035] In the application, the logical linkage organization between the structure parameters is realized by constructing the parameter control path and combining the multi-relationship identification between the structure nodes, the constraint identification ability for the structure control path is enhanced by using the cross relationship judgment mechanism between the standard clauses, the structure path screening has the semantic guidance feature by combining the index relationship of the task semantic label, the structure matching precision of the path screening is optimized by using the continuity judgment method of the closed path, the collaborative adaptability between the multiple structures is improved by using the response construction logic of the structure combination relationship, the linkage path with the continuity and logical attachment features is established, and the logical integrity and response coordination of the water conservancy design are improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The system flowchart of the application.

[0037] Figure 2 The flow chart of the parameter constraint construction module of the present application;

[0038] Figure 3 The flow chart of the design standard fusion module of the present application;

[0039] Figure 4 The flow chart of the task semantic attribution module of the present application;

[0040] Figure 5 The flow chart of the path screening closure module of the present application;

[0041] Figure 6 The flow chart of the structure combination response module of the present application. DETAILED DESCRIPTION

[0042] In order to make the objects, technical solutions and advantages of the present application clearer and more comprehensible, the present application will be further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0043] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0044] Please refer to Figure 1 The water conservancy design decision support system based on a knowledge graph comprises:

[0045] The parameter constraint construction module calls an engineering survey data interface, analyzes the engineering properties of various survey data and the role path in the design, calculates the linkage relationship between the rock strength and the lining thickness and establishes a parameter control path, calibrates the constraint transmission chain between the structure nodes, and obtains the parameter logic dependent structure;

[0046] The design standard fusion module uses the parameter logic dependent structure, analyzes the control boundary between the structure nodes and the design parameters, retrieves the constraint clauses of the structure parameters in the standard text, calculates the control combination between the controlled parameters and establishes a connection path, identifies the structure nodes that exist in the common control path, and obtains the index logic fusion density;

[0047] The task semantic attribution module analyzes the function distribution of the structural nodes in the task description based on the index logic fusion density, identifies the dominant parameter control path and matches the task semantic keywords, establishes the index relationship between the semantic labels and the master nodes, and obtains the master semantic identification label;

[0048] The path screening closure module analyzes the path parameter set associated with the master node in the atlas based on the master semantic identification label, calculates the continuous connection relationship between the parameters and screens the effective path set, and obtains the path closure structure record;

[0049] The structure combination response module analyzes the multi-structure combination relationship in the continuous path segment based on the path closure structure record, extracts the structural nodes including the dam body, diversion channel and discharge facility, calculates the geometric correspondence and control proportion between the structures, establishes the response connection path and records the path information, and obtains the combination response coordination information.

[0050] The parameter logic dependent structure includes the water head parameter control boundary, the seepage intensity linkage relationship and the structural node dependent path, the index logic fusion density is specifically the structural parameter control number distribution, the standard clause control level and the intersection identification between clauses, the master semantic identification label includes the function control label, the semantic master path identification and the task entry mapping sequence, the path closure structure record is specifically the structural path connection information, the boundary alignment matching order and the path number screening list, and the combination response coordination information includes the structural node response sequence, the combination control proportion relationship and the connection path index information.

[0051] Please refer to Figure 2 , the parameter constraint construction module includes:

[0052] The survey parameter identification submodule calls the engineering survey data interface, analyzes the engineering properties of various survey data and the role path in the design, including water head height, permeability coefficient and rock strength, judges the coupling trend of the change interval of the permeability coefficient at the stratum boundary and the change rate of the rock strength, establishes the engineering property matching table of hydrological permeability characteristics and rock mass structure response, and obtains the survey parameter response relationship comparison value;

[0053] After calling the engineering survey data interface and analyzing the engineering properties of the water head height of meters , the permeability coefficient of cm / s , and the rock strength of MPa , the specific implementation process is as follows: for the stratum boundary at the dam site, collect the permeability coefficient data at intervals of meters in the depth interval of meters to meters, obtain the data set , , , , ]( Simultaneously, rock strength data at corresponding depths were collected to obtain a dataset. , , , , ]( ), calculate the array of rates of change of the permeability coefficient [ , , , ] and rock stratum strength change rate array [ , , , Determine the sign of the product of two rate-of-change arrays at corresponding positions. If the product is negative, it is denoted as "negative correlation coupling," such as in depth. Permeability coefficient change rate at a depth of meters With the rate of change of rock layer strength The product is This was determined to be a negatively correlated coupling, and this negatively correlated coupling trend was then linked to the corresponding hydrological permeability characteristics. ( "and rock mass structure response" ( The data is recorded together, and all collection points are traversed to complete the matching. Finally, an engineering attribute matching table is established, which contains four fields: "stratum location", "permeability coefficient value", "rock strength value", and "coupling trend". Specified engineering nodes are extracted from this table, such as the depth of the area where the lining structure is located. Data at a distance of meters was used to obtain a comparison value for the response relationship of survey parameters.

[0054] The control relationship derivation submodule analyzes the stability of the response of rock layer strength changes to lining thickness based on the comparison values ​​of the survey parameter response relationship, and calculates the boundary coupling state between rock layer strength, water head height, and permeability coefficient using the following formula:

[0055] ;

[0056] Calculate the constraint strength value of the linkage structure, determine the direction of response transmission between the linkage structures, and generate the linkage parameter action coefficient;

[0057] in, This represents the constraint strength value of the linked structure. This is the normalized value of the rock stratum strength, obtained by reducing the original rock stratum strength data to its maximum value. is a normalized value of lining thickness, obtained by maximum reduction processing on original lining thickness data, is a normalized value of permeability coefficient, obtained by permeability test data reduction calculation, is a normalized value of water head height, obtained by partition water pressure monitoring data reduction processing, is a structure thickness boundary difference ratio value, obtained by dividing the maximum and minimum structure thickness difference by the average thickness;

[0058] According to the survey parameter response relationship value, when analyzing the rock strength ( ) response to the design lining thickness ( ), the boundary coupling state among the rock strength, water head height, and permeability coefficient is calculated, and the specific execution process is to call the formula to calculate the linkage structure constraint strength value, wherein is the linkage structure constraint strength value, is a normalized value of rock strength, is a normalized value of lining thickness, is a normalized value of permeability coefficient, is a normalized value of water head height, is a structure thickness boundary difference ratio value, and the calculation logic of the formula is that the numerator part quantifies the imbalance degree between the structure constraint and the hydrogeological constraint by calculating the absolute value of the difference between the product term of the rock and lining interaction and the root term of the hydrogeological condition influence, and the denominator part is a normalization factor that comprehensively considers the size variability of the structure itself and its normalized thickness. The formula quantifies the strength of the constraint transmission between structures under the action of multiple factors by defining the coupling relationship of geotechnical and hydrological parameters. The original data of each parameter is shown in Table 1:

[0059] Table 1 Original data table of survey parameters

[0060] ;

[0061] As shown in Table 1, first, the parameters are normalized and calculated. The normalized value of rock strength is , the normalized value of lining thickness is , the normalized value of permeability coefficient is , the normalized value of water head height is , and the structure thickness boundary difference ratio value is . The parameter values obtained by the above calculation are brought into the formula:

[0062] ;

[0063] The interconnected structural constraint strength value is a comprehensive parameter measuring the control and conduction strength between different engineering nodes in a hydraulic structural system under specific geological and hydrological conditions. A larger value indicates that a change in the parameter of one node in the interconnected structural pair will trigger a greater response from other related structural nodes; conversely, a smaller value indicates a limited impact. The core function of this parameter is to assess which locations in the structural system possess highly coupled and efficient critical paths, which can be used to construct more stable and responsive structural schemes. The results provide data support for subsequent path dependency construction and parameter optimization, serving as a preliminary quantitative basis for identifying strongly coupled node pairs in the constraint chain. The calculated interconnected structural constraint strength value... Compared with the preset intensity range, the intensity range is divided into: low influence zone ( ), moderate impact zone ), high-impact areas ), current calculation results Located in a low-impact zone, this result indicates that the constraint transmission between rock strength and lining thickness is weak. Based on this result, it is determined that the response transmission direction between the linked structures is dominated by hydrogeological conditions, ultimately generating the linkage parameter action coefficient.

[0064] The path dependency construction submodule analyzes the logical dependency order of the control path and establishes directional connections between nodes based on the linkage parameter action coefficient, records the parameter source information associated with the control path, identifies and marks the constraint transmission chain between structural nodes, and obtains the parameter logical dependency structure.

[0065] Based on the linkage parameter coefficient, When analyzing the logical dependency sequence of the control path, the nodes involved in the calculation are first identified, namely "rock strength", "lining thickness", "permeability coefficient" and "water head height", based on the constraint strength value of the linkage structure. Located in the low impact zone The judgment of "rock stratum strength" indicates that its direct control over "lining thickness" is relatively weak, while the square root term in the formula calculation... numerical value Much larger than the product term numerical value Based on this, directional connections between nodes are established. Specifically, a directed edge is created from the "permeability coefficient" node to the "lining thickness" node, and then a directed edge is created from the "water head height" node to the "lining thickness" node. Simultaneously, since rock strength remains a fundamental condition, a secondary directed edge is created from "rock strength" to "lining thickness," and each edge is assigned a weight. This weight value is the interaction coefficient of the linkage parameter generated in the previous step. For example, the weight of the path from "permeability coefficient" to "lining thickness" is set to... The weight of the path from "rock strength" to "lining thickness" is set as Then, the parameter source information associated with each control path is recorded, such as "permeability coefficient" derived from field water pressure test data and "rock strength" derived from drilling core test data. In this way, the core constraint transmission chain "hydrogeological conditions→lining structure" is identified and calibrated to obtain the parameter logical attachment structure.

[0066] Please refer to Figure 3 The design standard fusion module includes:

[0067] The clause attribution determination submodule obtains the parameter logical attachment structure, analyzes the control boundaries between the structure nodes and the design parameters, retrieves the constraint clauses of the structure parameters in the standard text, including the thickness of the protective layer, the width of the foundation, and the cross-section size, determines the control attribution relationship of the standard items corresponding to the structure nodes, identifies the parameter correspondence mode between the structure nodes and the design clauses, establishes the attribution distribution between the structure nodes and the standard clauses, and obtains the standard attribution distribution value.

[0068] After obtaining the parameter logical attachment structure, when analyzing the control boundaries between the "lining thickness" node and the design parameters, the relevant text of the hydraulic concrete structure design specification is retrieved, and the constraint clauses related to "protective layer thickness", "foundation width" and "cross-section size" are located, such as clause which stipulates the minimum protective layer thickness, and clause which stipulates the foundation width limit. Then, the control attribution relationship of the "lining thickness" node and these standard items is determined. Specifically, the parameter set of the "lining thickness" node is extracted, including "thickness value", "concrete grade", and "steel bar diameter". The frequency of these parameters in clauses and is counted, and the association ratio of "thickness value" with clause is calculated as , and the association ratio of "thickness value" with clause is calculated as Then, the attribution determination threshold is set, which is calculated by the difference between the average constraint strength of the node parameters and the boundary constraint value of the same parameter item in the standard clause. From the previous step, the average constraint strength of the "lining thickness" node is known to be , and the boundary constraint value of the thickness in clause is obtained from the specification as m , which is normalized to , and the threshold is calculated as The association ratio of the node parameters is compared with the attribution determination threshold , and because Lining Thickness node belongs to clause , and record the serial number of the connection in the home distribution as , repeat this process, and finally generate the home distribution between the structure node and the standard clause according to the number of home connections and the associated ratio weight, to obtain the standard home distribution value.

[0069] The parameter connection generation submodule calculates the control direction combination between the controlled parameters based on the standard home distribution value, and establishes the parameter connection path, combines the parameter connection paths between multiple structure nodes, calculates the connection situation of each group of parameters, and generates the parameter connection combination coefficient;

[0070] Based on the standard home distribution value, when calculating the control direction combination between the controlled parameters, it is identified that the Lining Thickness node belongs to clause , the Foundation Width node belongs to clause , through text analysis, it is found that clause and clause both refer to the clause about Permeability Resistance Grade , and accordingly a parameter connection path is established from Lining Thickness to Permeability Resistance Grade node and finally to Foundation Width node, at the same time, it is analyzed that another structure node Dam Slope belongs to clause , which is jointly constrained by the Foundation Bearing Capacity parameter with clause , therefore another connection path from Dam Slope to Foundation Width is established, combining the two paths forms a parameter connection path combination containing Lining Thickness, Foundation Width, and Dam Slope nodes, then the connection situation of this group of parameters is calculated, and the connection tightness is defined as the number of shared upstream constraint nodes, in this example, Lining Thickness and Foundation Width do not share upstream nodes, Dam Slope and Foundation Width share the Foundation Bearing Capacity node, so the connection tightness is , finally the connection tightness values of all path combinations are counted to generate the parameter connection combination coefficient.

[0071] The index cross calculation submodule calls the parameter connection combination coefficient, compares the cross situation of parameter control content in each standard clause, identifies the structure nodes that exist in the common control path and records the associated clause number, analyzes the multi-directional control relationship of the design standard in the graph structure, and obtains the index logical fusion density.

[0072] Call the parameter connection combination coefficient when comparing the cross situation of parameter control content in each standard clause, extract the parameter connection path generated in the last module, such as the Lining Thickness→Permeability Resistance Grade→Foundation Width path, and record its associated clause numbers as clause and clause and terms Simultaneously, another path is extracted: "Rock stratum strength → Foundation bearing capacity → Foundation width," and its associated clauses are clauses. Terms and Conditions and terms A comparison revealed that the "basic width" node at the end of both paths is associated with the clause. Therefore, the "Basic Width" node is identified as a structural node with a common control path, and its associated common clause number is recorded. and the upstream terms set { , Subsequently, the multi-directional control relationships of the design standards in the diagram structure were analyzed. It was found that the "foundation width" node received control inputs from two different directions: one from the hydraulic requirement of "permeability grade," and the other from the geotechnical requirement of "foundation bearing capacity." The number of such multi-directional control nodes was divided by the total number of all structural nodes in the diagram to calculate the value. Obtain the density of logical fusion of indicators.

[0073] Please see Figure 4 The task semantic attribution module includes:

[0074] The functional distribution extraction submodule analyzes the functional content of structural nodes in the task description based on the index logic fusion density, retrieves the functional parameters associated with the nodes, including storage capacity, flow rate regulation and outflow rhythm, analyzes the distribution of each parameter in the structural nodes, identifies the functional affiliation of nodes in the task scenario, and obtains the node functional distribution coefficient.

[0075] Based on the index logic fusion density is When analyzing the functional content of the "spillage facility" structural node in the task description "emergency flood discharge," the functional parameters associated with the "spillage facility" node were retrieved, and "gate opening," "weir crest elevation," and "channel roughness" were extracted. Then, the distribution of each parameter in the structural node was analyzed. Through design document analysis, the direct correlation between "gate opening" and the "emergency flood discharge" task was found to be... The "weir crest elevation" is primarily related to the "normal water storage" task, with a correlation degree of [missing information]. The correlation between "channel roughness" and the "long-term water conveyance efficiency" task is [correlation degree]. Based on this correlation distribution, the "water discharge facility" node is classified as a "direct control area" in the "emergency flood discharge" scenario and an "indirect influence area" in the "normal water storage" scenario. Finally, the correlation values ​​of each functional parameter are determined... , , After normalization, the result is calculated as follows: , , ] to obtain a node function distribution coefficient.

[0076] The parameter influence determination submodule calculates the influence direction of each structural parameter in the control path according to the node function distribution coefficient, compares the functional control differences of multiple structural nodes in the task path, identifies the dominant parameter and constructs a master parameter path, and obtains a dominant parameter influence path value.

[0077] According to the node function distribution coefficient , , ] when calculating the influence direction of each structural parameter in the control path, it is determined that the influence direction of the "gate opening" parameter (coefficient ) is positive, that is, increasing the opening can increase the flood discharge, and the influence direction of the "channel roughness" parameter (coefficient ) is negative, that is, increasing the roughness will reduce the flood discharge. Then, the functional control differences of the "flood discharge facility" node and the "upstream reservoir" node in the "emergency flood discharge" task path are compared, and it is found that the control coefficient of the "flood discharge facility" is much larger than the control coefficient of the "upstream reservoir". Accordingly, the "gate opening" is identified as the dominant parameter, and a path starting from the "operation instruction" node, pointing to the "gate opening" node, and then pointing to the "flood discharge flow" node is constructed as the master parameter path. The node function distribution coefficient values of each node on this path are weighted and summed to calculate , and the dominant parameter influence path value is obtained.

[0078] The semantic label mapping submodule calls the dominant parameter influence path value, identifies the task semantic keywords corresponding to the dominant parameter, establishes the index relationship between the semantic labels and the master nodes, constructs the atlas task entry mapping and records the semantic label binding information, and obtains the master semantic recognition label.

[0079] The dominant parameter influence path value is called when identifying the task semantic keywords corresponding to the dominant parameter "gate opening". From the task description "to reduce the downstream river water level to the safety water level within hours", the keywords "water level reduction" and "fast" are extracted, and the dominant parameter influence path value is compared with the preset semantic association threshold value. The threshold value is calculated according to the statistical average value of the master parameter influence value in the history similar engineering task and the standard deviation , which is set to the average value plus one standard deviation, that is . Since , it is determined that the "gate opening degree" is strongly associated with the keywords "fast" and "water level reduction", and then an index relationship between the semantic label "fast water level regulation" and the master control node "gate opening degree" is established, the mapping relationship is stored in the database, the atlas task entry mapping is constructed, and the binding information of the semantic label is recorded, including the creation time, the associated task number and the influence path value, and the master semantic recognition label is obtained.

[0080] Please refer to Figure 5 , the path screening closing module comprises:

[0081] The path parameter extraction submodule analyzes the path parameter set associated with the master node in the atlas based on the master semantic recognition label, retrieves the structural nodes related to the channel slope, the regulation and storage section and the cross section width, numbers each path segment and records the parameter index, and obtains the path structure parameter sequence;

[0082] Based on the master semantic recognition label "fast water level regulation", when analyzing the path parameter set associated with the master node "gate opening degree" in the atlas, a complete physical path is retrieved, which involves three structural nodes "upstream diversion channel", "gate control section" and "downstream spillway" in turn, and the design parameters of the three nodes are further retrieved, obtaining the parameter value related to "channel slope" , the parameter value related to "regulation and storage section" is length meters , the parameter value related to "cross section width" is meters , meters and meters at three nodes respectively, the two continuous path segments of the path are numbered, segment is "upstream diversion channel" to "gate control section", segment is "gate control section" to "downstream spillway", and the parameter index of each path segment is recorded, and finally the path structure parameter sequence is obtained.

[0083] The boundary connection determination submodule determines the continuous connection relationship between the structural segment parameters according to the path structure parameter sequence, determines the order and connectivity of the boundary intersection in the parameter transmission path, and uses the formula:

[0084] ;

[0085] The path continuous matching degree value is calculated, the continuous matching degree value is compared with the structural segment connection standard value, and the path structure alignment index is obtained;

[0086] Among them, This is the path continuity matching value, used to measure the sequential coherence of parameter connections between path structure segments. The number of path structure segments is obtained by counting the number of connections between structure nodes in each path. This is the index number of the current path structure segment, used to index the first... Parameters in the segment path This is the index number of the next segment after the current segment, used to construct the parameter difference between adjacent segments. For the first The normalized value of the segment structure path length is obtained by extracting the segment length and dividing it by the maximum segment length within the path. For the first The normalized value of the cross-sectional width difference of the segment structure path is calculated by the first... Section and the The difference in segment width is obtained by dividing by the maximum difference in width across the entire path. For the first The normalized value of the terrain variation range in the area where the route segment is located is obtained by comparing and normalizing the elevation differences between the start and end points of the nodes. For the first The normalized value of the density of water storage sections is obtained by statistically analyzing the number of water storage facilities per unit path length and then normalizing it. For the first The normalized value of the logical node code of a segment is obtained by normalizing the node logical sequence number. For the first The normalized value of the logical node encoding of the segment, and Similarly, it is obtained through normalized logical numbering;

[0087] Based on the path structure parameter sequence, when determining the continuous connection relationship between structure segment parameters, the formula is called. Calculate the path continuity matching degree value, where This represents the path continuity matching degree. This represents the number of path structure segments. This is the index number of the current path segment. For the first Normalized value of segment structure path length. For the first Normalized value of the cross-sectional width difference of the segment structure path. For the first The normalized value of the terrain variation range in the area where the route segment is located. For the first The normalized value of the density of the storage section. For the first The normalized value of the logical node encoding of the segment; the calculation logic of this formula lies in the numerator term. The degree of geometric abrupt change in the path's length and width is quantified, and the square root term in the denominator... This represents the buffering capacity of the terrain and water storage facilities on the water flow; the absolute value term. This represents the continuity of the logical order. The entire fraction evaluates the inconsistency of the connection between each segment. Finally, the overall degree of mismatch of the entire path is obtained by summing and averaging. The formula quantifies the physical and logical coherence of the path by integrating geometric, geological and logical parameters. The original data of the path segments are shown in Table 2.

[0088] Table 2 Path Structure Segment Parameter Table

[0089] ;

[0090] As shown in Table 2, in order to substitute the physical dimensional data in Table 2 into the formula To perform dimensionless calculations, it is necessary to first establish a normalized baseline value based on the full path data and establish a mapping relationship between the original parameters and formula symbols. The specific extraction and definition process is as follows:

[0091] First, extract the normalized baseline value (denominator) of the entire path: Traverse all path segment data in Table 2 and identify the maximum segment length of the entire path. for (Taken from section 2); Calculate the absolute value of the difference in cross-sectional width for each section ( , Extract the maximum width difference for Extract the overall elevation difference along the path. Starting point and the end point The difference, that is ; Calculate the storage density of each section (section 1 is...) Section 2 is Extract the maximum storage density. for Extract the maximum logical code. for (Taken from the next logical code).

[0092] Second, define the computational mapping (numerator) between formula symbols and the data in Table 2: In the formula Corresponding to "Path Segment Length" in Table 2 "and The ratio; in the formula The absolute value of the difference in width between the starting and ending sections corresponds to the value in Table 2. The ratio; in the formula The absolute value of the starting / ending elevation difference corresponds to the value in Table 2. The ratio; in the formula Corresponding to the "Number of Storage and Regulation Facilities" in Table 2 Divide by length The obtained density value and The ratio; in the formula Corresponding to "Logical Encoding" in Table 2 The ratio;

[0093] Number of path segments First, the parameters are calculated and normalized to determine the maximum segment length of the entire path. ( ), maximum width difference of the entire path ( Maximum elevation difference along the entire path ( Maximum storage density along the entire path Maximum logical encoding of the entire path For segment 1 ( ): , , , , , For segment 2 ( ): , , , , , Substitute the parameter values ​​into the formula:

[0094] ;

[0095] ;

[0096] ;

[0097] The path continuity matching value is an indicator used to measure the stability of parameter continuous transmission in a structural path. Essentially, it assesses whether the connection and coherence between structural nodes in the graph path conforms to the design logic. A smaller value indicates a more coordinated match between structural segments in terms of structural parameters, logical relationships, and physical conditions, resulting in better path continuity and suitability for closed path identification and effective path selection. Conversely, a larger value indicates structural abrupt changes, logical jumps, or sudden drops in regulation performance, which can easily lead to path closure failure or response chain interruption. This parameter serves as a crucial threshold for determining path validity, providing a structural closure basis for subsequent structural combination response and task semantic mapping. A structural segment connection standard value of 0.3 is set; this standard value is obtained by analyzing similar paths from 10 successfully operating water conservancy projects. The value was calculated using the 80th percentile of the result. The calculated result of 0.433 was compared to the standard value of 0.3. The path structure alignment index was found to be "not satisfied".

[0098] The effective path filtering submodule identifies broken and mismatched nodes in the path based on the path structure alignment index and records the path number. It removes path segments with missing synchronization control relationships, filters the effective path set, establishes a continuous verification chain, and obtains the path closure structure record.

[0099] Based on the path structure alignment index being "not satisfied" and the path continuity matching value being 0.433, when identifying nodes with breaks and boundary mismatches in the path, backtracking the calculation process revealed that the calculation item for the second path segment... Its value is much higher than 0.205 of the first path segment. Based on this, it is determined that there is a significant boundary mismatch at the connection between the "gate control section" and the "downstream spillway". The path number "Path-001" and the mismatch node "downstream spillway" are recorded. Then, the path is removed from the set of valid paths. Next, the system continues to analyze the next candidate path and repeats the boundary connectivity determination process until all paths are analyzed. Paths with a continuous matching degree value of less than 0.3 are selected. A continuous verification chain is established for these selected paths to obtain the path closure structure record.

[0100] Please see Figure 6 The structural combination response module includes:

[0101] The structural relationship extraction submodule analyzes the multi-structure combination relationship within a continuous path segment based on the path closure structure record, extracts structural nodes including dam body, diversion channel and water discharge facility, counts the combination distribution of structural nodes, identifies structural combination relationship, and obtains structural node combination distribution table;

[0102] Based on the path closure structure record, when analyzing the multi-structure combination relationship within the continuous path segment of a certain effective path "Path-002", the structural nodes arranged sequentially on this path are extracted, namely "dam body", "diversion channel" and "spillage facility". The combination distribution of these three nodes is statistically analyzed, revealing that they exist in all... The same appeared in all the valid paths Therefore, the "dam body-diversion channel-discharge facility" is identified as a high-frequency structural combination relationship, and this combination relationship is associated with its frequency of occurrence. Record all valid paths and all node combinations together, and finally generate a structured node combination distribution table containing fields such as "nodes within the combination", "frequency of occurrence of the combination", and "path number".

[0103] The parameter linkage judgment submodule calculates the corresponding ratio between the dam body inclination angle and the cross-sectional shape of the diversion channel based on the structural node combination distribution table, analyzes the influence direction of the inclination angle change on the geometric characteristics of the diversion cross-section, determines the control parameters of the channel cross-section change on the flow capacity of the water discharge facility, and obtains the parameter linkage response value.

[0104] Based on the structural node combination distribution table, when calculating the corresponding ratio between the dam body inclination angle and the diversion channel cross-sectional shape in the "dam body-diversion channel" combination, the following parameters are extracted: A set of successful case data, in which the dam body inclination angles are respectively [ , , , , The corresponding cross-sectional shapes of the flow guiding channels (expressed as width-to-depth ratio) are as follows: , , , , ] Calculate the covariance of the two sets of data. The standard deviation of the dam body inclination angle is The standard deviation of the width-to-depth ratio is The correlation coefficient was obtained as Based on this, it is determined that as the dam inclination angle increases, the width-to-depth ratio of the diversion channel also increases, with the influence direction being positive. Next, it is determined that the control parameter for the flow capacity of the spillway facility based on the change in the width-to-depth ratio of the channel cross-section is the "inlet velocity." The calculation is performed when the width-to-depth ratio changes from... Increase to At that time, the inlet flow rate was from ( Reduced to ( Finally, this linkage relationship is quantified to obtain the parameter linkage response value.

[0105] The response path generation submodule calls the parameter linkage response value, establishes the response connection path between the diversion node and the discharge node, identifies the parameter transmission order between nodes in the path, analyzes the stability of parameter control transmission between nodes, identifies the structural combination that forms the linkage response and records the path information, and obtains the combined response coordination information.

[0106] The response value of the calling parameter linkage is responded, when the response connection path between the diversion node and the drainage node is established, the linkage relationship of the "inlet flow velocity of the drainage facility" based on the "width-depth ratio of the diversion channel" determined in the last module is controlled, a directed edge from the "diversion channel" node to the "drainage facility" node is created, and the response type is identified as "negative correlation". Then, the parameter transmission sequence between the nodes in the path is identified as: the engineer adjusts the "dam body inclination" parameter, the change is transmitted to the "width-depth ratio of the diversion channel" parameter through the "dam body-diversion channel" combination relationship, and then transmitted to the "inlet flow velocity of the drainage facility" parameter through the "diversion-drainage" linkage relationship. Then, the stability of the parameter control transmission between the nodes is analyzed, the transmission function from the "dam body inclination" to the "inlet flow velocity" is calculated according to the historical data, the first derivative is , the fluctuation is less than , and it is determined that the transmission is stable. Finally, the "dam body-diversion channel-drainage facility" structure combination is identified as a structure combination forming a linkage response, and the complete response path information is recorded to obtain the combination response coordination information.

[0107] The above is only a preferred embodiment of the present application, and does not limit the present application in other forms. Any skilled person in the art can modify or change the above disclosed technical content to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application technical solution.

Claims

1. A knowledge graph-based hydraulic engineering design decision support system, characterized in that, The system includes: The parameter constraint construction module calls the engineering survey data interface to analyze the engineering attributes of various survey data and their role paths in the design, calculates the linkage relationship between rock layer strength and lining thickness and establishes parameter control paths, calibrates the constraint transmission chain between structural nodes, and obtains the parameter logical dependent structure. The design standard fusion module utilizes the parameter logic dependency structure to analyze the control boundary between structural nodes and design parameters, retrieve constraint clauses for structural parameters in the standard text, calculate control combinations between controlled parameters and establish connection paths, identify structural nodes with common control paths, and obtain the index logic fusion density. Based on the logical fusion density of the indicators, the task semantic attribution module analyzes the functional distribution of structural nodes in the task description, identifies the dominant parameter control path and matches task semantic keywords, establishes the index relationship between semantic tags and master control nodes, and obtains master control semantic identification tags. The path filtering and closure module analyzes the set of path parameters associated with the master control node in the graph based on the master control semantic recognition tag, calculates the continuous connection relationship between the parameters and filters the set of valid paths to obtain the path closure structure record; The parameter logical dependency structure includes the head parameter control boundary, the permeability intensity linkage relationship, and the structural node dependency path. The indicator logical fusion density specifically includes the distribution of structural parameter control numbers, the standard clause control hierarchy, and the boundary intersection identifier between clauses. The master control semantic recognition label includes the function control label, the semantic master control path identifier, and the task entry mapping sequence. The path closure structure record specifically includes the structural path connection information, the boundary alignment matching order, and the path number filtering list.

2. The knowledge graph-based hydraulic design decision support system according to claim 1, characterized in that, The parameter constraint construction module includes: The survey parameter identification submodule calls the engineering survey data interface to analyze the engineering attributes of various survey data and their role in the design, including water head height, permeability coefficient and rock strength. It determines the coupling trend between the variation range of permeability coefficient at the stratigraphic boundary and the rate of change of rock strength, establishes an engineering attribute matching table between hydrological permeability characteristics and rock mass structure response, and obtains the reference value of the survey parameter response relationship. The control relationship derivation submodule analyzes the stability of the response of rock layer strength changes to lining thickness based on the reference values ​​of the survey parameter response relationship, and calculates the boundary coupling state between rock layer strength, hydraulic head, and permeability coefficient using the following formula: ； Calculate the constraint strength value of the linkage structure, determine the direction of response transmission between the linkage structures, and generate the linkage parameter action coefficient; in, This represents the constraint strength value of the linked structure. This is the normalized value of the rock stratum strength. This is the normalized value for the lining thickness. This is the normalized value of the permeability coefficient. This is the normalized value of the water head height. This represents the ratio of the difference in structural thickness at the boundary. The path dependency construction submodule analyzes the logical dependency order of the control path and establishes directional connections between nodes based on the linkage parameter action coefficient, records the parameter source information associated with the control path, identifies and marks the constraint transmission chain between structural nodes, and obtains the parameter logical dependency structure.

3. The knowledge graph-based hydraulic design decision support system according to claim 2, characterized in that, The design standard integration module includes: The clause attribution determination submodule obtains the parameter logical dependency structure, analyzes the control boundary between structural nodes and design parameters, retrieves the constraint clauses of structural parameters in the standard text, including protective layer thickness, foundation width and cross-sectional dimensions, determines the control attribution relationship of the standard item corresponding to the structural node, identifies the parameter correspondence between the structural node and the design clause, establishes the attribution distribution between the structural node and the standard clause, and obtains the standard attribution distribution value. The parameter connection generation submodule calculates the control direction combination between controlled parameters and establishes parameter connection paths based on the standard attribution distribution value, combines the parameter connection paths between multiple structural nodes, calculates the connection status of each group of parameters, and generates parameter connection combination coefficients. The indicator cross-calculation submodule calls the parameter connection combination coefficient, compares the cross-control content of parameters in each standard clause, identifies structural nodes with common control paths and records the associated clause number, analyzes the multi-directional control relationship of design standards in the graph structure, and obtains the indicator logic fusion density.

4. The knowledge graph-based hydraulic design decision support system according to claim 3, characterized in that, The process of establishing the attribution distribution between structural nodes and standard clauses is as follows: After identifying the control attribution relationship of the standard items corresponding to the structural nodes, extract the set of parameters related to the protective layer thickness, the foundation width and the cross-sectional dimensions for each structural node. Based on the frequency of occurrence and reference position of each parameter in the standard clause text, calculate the association ratio between the node parameters and the clause constraint content. Set the attribution determination threshold according to the distribution range of the association ratio. The attribution determination threshold is calculated by the difference between the average constraint strength of the node parameters and the boundary constraint value of the same parameter item in the standard clause. During the attribution calculation process, the association ratio of each node parameter is determined to be higher than the attribution determination threshold. If it is higher, the target node is established with the corresponding standard clause, and the sequence number of the target connection in the attribution distribution is recorded. Based on the number of attribution connections between each node and the standard clause and the association ratio weight, the attribution distribution between the structural node and the standard clause is generated, and the standard attribution distribution value is obtained.

5. The knowledge graph-based hydraulic design decision support system according to claim 3, characterized in that, The task semantic attribution module includes: The functional distribution extraction submodule analyzes the functional content of structural nodes in the task description based on the logical fusion density of the indicators, retrieves the functional parameters associated with the nodes, including storage capacity, flow rate regulation and outflow rhythm, analyzes the distribution of each parameter in the structural nodes, identifies the functional affiliation of nodes in the task scenario, and obtains the node functional distribution coefficient. The parameter influence determination submodule calculates the influence direction of each structural parameter in the control path based on the node function distribution coefficient, compares the functional control differences of multiple structural nodes in the task path, identifies the dominant parameter and constructs the main control parameter path, and obtains the dominant parameter influence path value. The semantic tag mapping submodule calls the dominant parameter to influence the path value, identifies the task semantic keywords corresponding to the dominant parameter, establishes the index relationship between semantic tags and master nodes, constructs the graph task entry mapping and records the semantic tag binding information, and obtains the master semantic recognition tag.

6. The knowledge graph-based hydraulic design decision support system according to claim 5, characterized in that, The path filtering closure module includes: The path parameter extraction submodule analyzes the set of path parameters associated with the main control node in the map based on the main control semantic recognition tag, retrieves structural nodes involving channel slope, water storage section and cross-sectional width, numbers each path segment and records the parameter index to obtain the path structure parameter sequence; The boundary connectivity determination submodule determines the continuous connection relationship between structural segment parameters based on the path structure parameter sequence, determines the order and connectivity of boundary intersections in the parameter transmission path, calculates the path continuity matching degree value, compares the continuity matching degree value with the structural segment connection standard value, and obtains the path structure alignment index. The effective path filtering submodule identifies nodes with breaks and mismatched boundaries in the path based on the path structure alignment index and records the path number. It then removes path segments with missing synchronization control relationships, filters the set of effective paths, establishes a continuous verification chain, and obtains the path closure structure record.

7. The knowledge graph-based hydraulic design decision support system according to claim 1, characterized in that, The system also includes: Based on the path closure structure record, the structural combination response module analyzes the multi-structure combination relationship in the continuous path segment, extracts structural nodes including dam body, diversion channel and water discharge facility, calculates the geometric correspondence and control ratio between structures, establishes response connection path and records path information, and obtains combined response coordination information. The combined response coordination information includes structural node response sequences, combined control ratio relationships, and connection path index information.

8. The knowledge graph-based hydraulic design decision support system according to claim 7, characterized in that, The structural combination response module includes: The structural relationship extraction submodule analyzes the multi-structure combination relationship within a continuous path segment based on the path closure structure record, extracts structural nodes including dam body, diversion channel and water discharge facility, counts the combination distribution of structural nodes, identifies structural combination relationships, and obtains a structural node combination distribution table. The parameter linkage determination submodule calculates the corresponding ratio between the dam body inclination angle and the cross-sectional shape of the diversion channel according to the structural node combination distribution table, analyzes the influence direction of the inclination angle change on the geometric characteristics of the diversion cross-section, determines the control parameters of the channel cross-section change on the flow capacity of the water discharge facility, and obtains the parameter linkage response value. The response path generation submodule calls the parameter linkage response value to establish a response connection path between the diversion node and the discharge node, identifies the parameter transmission order between nodes in the path, analyzes the stability of parameter control transmission between nodes, identifies the structural combination that forms the linkage response and records the path information, and obtains the combined response coordination information.

Images