Knowledge graph construction method and system based on reservoir group scheduling boundary and constraint library

Abstract

The invention belongs to the related technical field of intelligent water conservancy construction, and discloses a knowledge graph construction method and system based on a reservoir group scheduling boundary and a constraint library. The method comprises the following steps: S1, obtaining different hydraulic engineering characteristics, and scheduling boundaries and constraints of operation control parameters of a reservoir in different scheduling stages and different scenes; s2, constructing a water conservancy project feature sub-map, a scheduling stage sub-map and a scheduling scene sub-map; and S3, merging the same entity nodes in the hydraulic engineering feature sub-atlas, the scheduling stage sub-atlas and the scheduling scene sub-atlas to realize fusion of each sub-atlas, thereby obtaining the required knowledge atlas. According to the invention, unified visual display of scheduling boundaries and constraint conditions and intelligent setting of reservoir managers for the boundaries and the constraint conditions are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field related to the construction of intelligent water conservancy, and more specifically, relates to a method and system for constructing a knowledge graph based on the scheduling boundary of a reservoir group and a constraint library. Background Art

[0002] The construction of intelligent water conservancy is a key path to promote the high-quality development of water conservancy in the new stage and is of great significance in optimizing the allocation of water resources and enhancing flood control and disaster reduction capabilities. In this context, improving the intelligent level of reservoir scheduling is an inevitable choice for the high-quality development of reservoir scheduling. The determination of the scheduling boundary and constraint conditions of a reservoir group is the basis for realizing the scheduling of a reservoir group. At present, there are no effective means for extraction, integration, and intelligent generation. Usually, it is statically set according to the basic parameters of the reservoir and the empirical knowledge of schedulers. This traditional method cannot meet the requirements of timeliness and overall situation of reservoir intelligent scheduling.

[0003] The knowledge graph technology is essentially a very large semantic network that integrates various heterogeneous information sources to represent the knowledge in certain specific fields. Its graphical structure not only promotes the interconnection between data but also strengthens the machine's understanding and processing of information, becoming an important cornerstone of modern artificial intelligence research. At present, the application of knowledge graph technology in the water conservancy field is gradually increasing, but it mainly focuses on extracting entities and relationships from water conservancy-related text datasets using natural language processing technology, often ignoring the key role of the schema layer of the knowledge graph and not conducting a detailed architecture design in the schema layer according to actual business requirements.

[0004] With the growth of the number of reservoirs and the refinement of scheduling periods, the dimension of decision variables will also increase sharply. The traditional method of statically setting the scheduling boundary and constraints manually is difficult to comprehensively consider various influencing factors, resulting in a relatively wide range of set scheduling boundaries and constraints, making it difficult for the optimization solution algorithm to quickly converge to a feasible solution. In addition, there is currently no method for effectively extracting the scheduling boundary and constraints of a reservoir group from multiple heterogeneous data sources; how to effectively integrate and display the extracted different types of scheduling boundary and constraint data is the key to constructing a reservoir group scheduling boundary and constraint library. Although knowledge graph technology has significant advantages in data organization and integration, intelligent recommendation, etc., its application in the water conservancy field is not sufficient, and the graph architecture is not clear enough, resulting in performance bottlenecks in processing complex queries and reasoning tasks. Summary of the Invention

[0005] In view of the above defects or improvement requirements of the prior art, the present invention provides a method and system for constructing a knowledge graph based on the scheduling boundary of a reservoir group and a constraint library, which solves the problem that the extraction and integration of heterogeneous data sources of various scheduling boundaries and constraints cannot be achieved in the prior art.

[0006] To achieve the above object, according to one aspect of the present invention, there is provided a method for constructing a knowledge graph based on reservoir group scheduling boundaries and a constraint library, the method comprising the following steps:

[0007] S1 Parse the unstructured data of the reservoir operation regulations and the historical operation data of the reservoir to obtain different water conservancy project characteristics, and the scheduling boundaries and constraints of the operation control parameters of the reservoir in different scheduling stages and different scenarios;

[0008] S2 Use the obtained different water conservancy project characteristics, and the scheduling boundaries and constraints of the water level, discharge flow, and output of the reservoir in different scheduling stages and different scenarios to construct their respective corresponding sub-graphs, including a water conservancy project characteristics sub-graph, a scheduling stage sub-graph, and a scheduling scenario sub-graph. Each sub-graph includes multiple entity nodes, and each entity node contains the entity attributes and entity content of the entity node. The entity nodes are connected through the relationship attributes between the entity nodes;

[0009] S3 Merge the same entity nodes in the water conservancy project characteristics sub-graph, the scheduling stage sub-graph, and the scheduling scenario sub-graph to realize the fusion of each sub-graph, thereby obtaining the required knowledge graph.

[0010] Further preferably, the large language model is used to parse the reservoir operation regulations. The input of the large language model is: prompt sentence + reservoir operation regulation text. The prompt sentence adopts the structure of the CO-STAR framework, and the output is the water conservancy project characteristics.

[0011] Further preferably, the sliding window statistical analysis method is used to parse the historical operation data of the reservoir.

[0012] Further preferably, when constructing the sub-graph, first convert the different water conservancy project characteristics, and the scheduling boundaries and constraints of the water level, discharge flow, and output of the reservoir in different scheduling stages and different scenarios into triples. The triples include entity node A, entity node B, and the relationship attribute between entity nodes A and B. The entity content in entity nodes A and B is displayed on the entity nodes of the sub-graph, and the entity attributes are not displayed.

[0013] Further preferably, the entity attributes in the water conservancy project characteristics sub-graph include the name of the water conservancy project, the water conservancy project grade value, the function of the water conservancy project, the check flood level value, the design flood level value, the flood control limit water level value, the low water level value for drawdown during the dry period, the installed capacity value, the unit type, and the unit discharge capacity value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, characteristic water levels, and generator set characteristics.

[0014] Further preferably, the entity attributes in the sub-graph of the scheduling stage include the name of the water conservancy project, the drawdown period, the flood season, the impounding period, the water level value, the discharge flow value, and the output value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include the characteristics of the water conservancy project, the scheduling stage, the lower historical water level, the upper historical water level, the lower historical flow, the upper historical flow, the lower historical output, the upper historical output, the lower regulated water level, the upper regulated water level, the lower regulated flow, the upper regulated flow, the lower regulated output, and the upper regulated output.

[0015] Further preferably, the entity attributes in the sub-graph of the characteristic parameters of the water conservancy project include the name of the water conservancy project, the drawdown period, the flood season, the impounding period, a certain month, a certain ten-day period, the water level value, the discharge flow value, the output value, and the condition value; the entity content includes the name or value corresponding to each entity attribute; the relationship attributes include the characteristics of the water conservancy project, the scheduling stage, the scheduling cycle, the boundaries and constraints, and the scheduling conditions.

[0016] Further preferably, the reservoir group parameters include the reservoir water level, the reservoir discharge flow, and the reservoir output.

[0017] According to another aspect of the present invention, there is provided a system for constructing a knowledge graph by using the knowledge graph construction method based on the reservoir group scheduling boundaries and constraint library described above. The system includes a data acquisition module, a sub-graph construction module, and a knowledge graph fusion module, wherein:

[0018] The data acquisition module is used to acquire the scheduling boundaries and constraints of different water conservancy project characteristics and the operation control parameters of the reservoir in different scheduling stages and different scenarios, and convert the acquired data into corresponding triples;

[0019] The knowledge graph construction module is used to construct a sub-graph of water conservancy project characteristics, a sub-graph of the scheduling stage, and a sub-graph of the scheduling scenario;

[0020] The knowledge graph fusion module is used to fuse the sub-graphs to form a final knowledge graph.

[0021] According to yet another aspect of the present invention, there is provided a system for constructing a knowledge graph by using the knowledge graph construction method based on the reservoir group scheduling boundaries and constraint library. The system includes an executor, which is used to execute the method for constructing a knowledge graph by using the above-mentioned knowledge graph construction method based on the reservoir group scheduling boundaries and constraint library.

[0022] Generally speaking, compared with the prior art, the above technical solutions conceived by the present invention have the following beneficial effects:

[0023] 1. The present invention constructs multiple sub-graphs by utilizing different water conservancy project characteristics, the scheduling boundaries and constraints of the operation control parameters of the reservoir in different scheduling stages and different scenarios, and then fuses the sub-graphs to obtain the required knowledge graph. This method can dynamically adjust the scheduling boundaries and constraints according to specific scheduling requirements and environmental conditions, ensuring the accuracy and real-time nature of the reservoir group scheduling, and solving the problems in the prior art such as the inability to efficiently extract and integrate information related to reservoir group scheduling, the overly broad setting of scheduling boundaries, and the performance bottleneck caused by the unclear knowledge graph architecture.

[0024] 2. The large language model for text processing and the statistical analysis method based on a sliding window of the present invention respectively extract the scheduling boundaries and constraint information from the reservoir scheduling regulations and historical scheduling processes. The range of the scheduling boundaries and constraints extracted is relatively narrow compared with the traditional method, which is beneficial to the optimization solution algorithm to quickly converge to a feasible solution. On this basis, sub-graphs are constructed and knowledge fusion is carried out to complete the intelligent construction of the reservoir scheduling boundary and constraint library, which helps to uniformly visualize the display of the scheduling boundaries and constraint conditions and the intelligent setting of the boundary and constraint conditions by reservoir managers.

[0025] 3. Aiming at the problems of the organization and integration, knowledge representation, and visualization of different types of scheduling boundary and constraint data, the present invention constructs a knowledge graph of reservoir group scheduling boundaries and constraints, effectively revealing the interconnection and hierarchical structure between different types of scheduling boundary and constraint data, which helps to uniformly visualize the display of the scheduling boundaries and constraint conditions and the intelligent setting of the boundary and constraint conditions by reservoir managers. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is a flowchart of a method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library constructed according to a preferred embodiment of the present invention;

[0027] Figure 2 is a schematic diagram of the prompt engineering strategy - CO-STAR framework when constructing a large language model for text processing according to a preferred embodiment of the present invention;

[0028] Figure 3 is the extraction result of the flood season water level constraint of the Three Gorges Reservoir obtained by the statistical analysis method based on a sliding window constructed according to a preferred embodiment of the present invention;

[0029] Figure 4 is a partial view of the sub-graph of the scheduling stage of the Three Gorges Reservoir constructed according to a preferred embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] To make the objectives, technical solutions, and advantages of the present invention more clear and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0031] As Figure 1 shown, a method for constructing a knowledge graph based on the reservoir group scheduling boundary and constraint library specifically includes the following steps:

[0032] S1 Parse the unstructured data of the reservoir operation regulations and the historical operation data of the reservoir to obtain different water conservancy project characteristics, scheduling boundaries, and constraints of the operation control parameters of the reservoir in different scheduling stages and different scenarios. The operation control parameters include reservoir water level, reservoir discharge flow, and reservoir output.

[0033] A. As Figure 2 shown, a large language model is used to parse the unstructured data of the reservoir operation regulations. The input of this large language model is: prompt sentence + reservoir operation regulation text. The prompt sentence adopts the structure of the CO-STAR framework, and the output is the water conservancy project characteristics. That is, the CO-STAR framework is used to build the structure of the large model prompt. When constructing the input instructions of the CO-STAR framework, the following 6 aspects are mainly considered:

[0034] (1) Context: Background information related to the given task, which helps the large model understand the specific scenario being discussed.

[0035] (2) Objective: Define the task that the large model needs to execute, which helps the large model focus its response on completing the specific task.

[0036] (3) Style: Specify the writing style used by the large model, which helps guide the large model to give a response using words and methods that meet the requirements.

[0037] (4) Tone: Set the attitude of the large model's response, which helps the large model's response conform to the required emotion or mood.

[0038] (5) Audience: Determine the target audience of the response to ensure that the large model's response is appropriate and understandable.

[0039] (6) Response: Provide the format of the response to ensure that the output of the large model conforms to the format of the downstream task.

[0040] B. The sliding window statistical analysis method is used to parse the historical operation data of the reservoir

[0041] The statistical analysis method based on a sliding window includes the following steps:

[0042] (1) Resample from historical data using a sliding window to obtain sample data.

[0043] (2) Calculate the statistics of each sample: maximum value, average value, minimum value. The calculation formulas for each statistic are as follows:

[0044]

[0045] In the formula: Xmax, Xmin are the sets of the maximum value, average value, and minimum value of the sample data obtained from the sliding window based on the original data. i represents the order of the sliding window, k represents the length of the slider, n represents the number of sliders included in each sliding window, T represents the number of years included in the historical data, ximax, ximin are the maximum value, average value, and minimum value of the sample data in the i-th window respectively, xi1,…,xi+n-1k are the sample data in the i-th window, and l represents the order of the data within the slider.

[0046] (3) According to the sampling distribution of the obtained statistics, perform interval estimation on each statistic, and calculate the confidence interval using the t-distribution. The calculation formula is:

[0047]

[0048] In the formula: are the average values of the maximum value, average value, and minimum value samples respectively, s max 、 s min are the standard deviations of the maximum value, average value, and minimum value samples respectively; t α / 2 is the critical value at the α / 2 positions of the upper and lower confidence levels of the t-distribution; T - n is the sample size.

[0049] Finally, take the range between the minimum confidence lower limit and the maximum confidence upper limit of the water level, discharge flow, and output constraint of each ten-day period of the reservoir as the extraction result.

[0050] S2 constructs their respective corresponding sub-graphs using the obtained different water conservancy project characteristics, the scheduling boundaries and constraints of the water level, discharge flow, and output of the reservoir in different scheduling stages and different scenarios, including the water conservancy project characteristics sub-graph, the scheduling stage sub-graph, and the scheduling scenario sub-graph. Each sub-graph includes multiple entity nodes, and each entity node contains the entity attributes and entity content of the entity node. The entity nodes are connected through the relationship attributes between the entity nodes.

[0051] When constructing the sub-graph, first convert the different water conservancy project characteristics, the water levels, discharge flows, and power generation capacities of the reservoir at different operation stages and under different scenarios, as well as the scheduling boundaries and constraints, into triples. A triple includes an entity node A, an entity node B, and the relationship attribute between entity nodes A and B. The entity content in entity nodes A and B is displayed on the entity nodes of the sub-graph, while the entity attributes are not displayed.

[0052] The entity attributes in the water conservancy project characteristics sub-graph include the name of the water conservancy project, the water conservancy project grade value, the function of the water conservancy project, the check flood level value, the design flood level value, the flood control limit water level value, the low water level for drawdown during the dry period value, the installed capacity value, the unit type, and the unit discharge capacity value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, characteristic water levels, and generator set characteristics.

[0053] The entity attributes in the operation stage sub-graph include the name of the water conservancy project, the drawdown period, the flood season, the impoundment period, the water level value, the discharge flow value, and the power generation capacity value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, operation stages, the lower bound of historical water levels, the upper bound of historical water levels, the lower bound of historical discharge flows, the upper bound of historical discharge flows, the lower bound of historical power generation capacities, the upper bound of historical power generation capacities, the lower bound of the regulated water level, the upper bound of the regulated water level, the lower bound of the regulated discharge flow, the upper bound of the regulated discharge flow, the lower bound of the regulated power generation capacity, and the upper bound of the regulated power generation capacity.

[0054] The entity attributes in the operation scenario sub-graph include the name of the water conservancy project, the drawdown period, the flood season, the impoundment period, a certain month, a certain ten-day period, the water level value, the discharge flow value, the power generation capacity value, and the condition value; the entity content includes the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, operation stages, operation cycles, boundaries and constraints, and operation conditions.

[0055] S3 merges the same entity nodes in the water conservancy project characteristics sub-graph, the operation stage sub-graph, and the operation scenario sub-graph to achieve the integration of each sub-graph, thereby obtaining the required knowledge graph.

[0056] The present invention will be further described below with specific embodiments.

[0057] Taking the Three Gorges Reservoir as an example, the intelligent extraction and construction method for the scheduling boundaries and constraints of reservoir groups based on the knowledge graph provided by the present invention will be elaborated. The basic data sources are the "Three Gorges (Normal Operation Period) - Gezhouba Water Control Project Cascade Scheduling Regulations (2019 Revision)" and the water level, discharge flow, and power generation capacity operation data from 2011 to 2021. The specific steps are as follows:

[0058] (1) Use a large language model for text processing to parse the unstructured data of the reservoir operation regulations, obtain the operation control requirements such as water levels and discharge flows of the reservoir under different operation stages and different water and rainfall scenarios. Examples of different operation scenario data extracted by the large language model using the CO-STAR prompt strategy are as follows:

[0059] Operation stage: Flood season;

[0060] Operation cycle: June 11 - August 20;

[0061] Condition entity 1: Real-time Three Gorges reservoir inflow;

[0062] Condition value 1: Less than 30000m 3 / s;

[0063] Condition entity 2: Forecasted Three Gorges reservoir inflow for the next three days;

[0064] Condition value 2: All not greater than 32000m 3 / s;

[0065] Condition entity 3: Water level at Shashi Station;

[0066] Condition value 3: Below 41m;

[0067] Condition entity 4: Water level at Chenglingji (Lianhuatang) Station;

[0068] Condition value 4: Below 30.5m

[0069] Operation boundary and constraint: The water level of the Three Gorges Reservoir is less than or equal to 146.5m.

[0070] (2) Use the statistical analysis method based on a sliding window to extract the operation range of the reservoir's water level, discharge flow, and output power in different pentads from the historical operation data of the reservoir, and obtain the complete operation boundary and constraint data. The extraction results of the flood season water level constraints of the Three Gorges Reservoir are as Figure 3 shown.

[0071] (3) According to the different characteristics of the operation boundary and constraint data, conduct detailed architecture design and clear ontology definition in the knowledge graph schema layer respectively.

[0072] Entities of the Three Gorges Reservoir's water conservancy project characteristics include the name of the water conservancy project, the value of the water conservancy project grade, the function of the water conservancy project, the value of the check flood level, the value of the design flood level, the flood control limit water level value, the low water level value for drawdown during the dry period, the installed capacity value, the type of generating units, the value of the discharge capacity of the generating units, etc. Relationship attributes include water conservancy project characteristics, characteristic water levels, and generating unit characteristics.

[0073] The entity attributes in the Three Gorges Reservoir operation stage include the name of the water conservancy project, the drawdown period, the flood season, the impoundment period, water level values, discharge values, and power generation values, etc. The relationship attributes include the characteristics of the water conservancy project, the operation stage, the lower bound of historical water levels, the upper bound of historical water levels, the lower bound of historical discharges, the upper bound of historical discharges, the lower bound of historical power generations, the upper bound of historical power generations, the lower bound of regulated water levels, the upper bound of regulated water levels, the lower bound of regulated discharges, the upper bound of regulated discharges, the lower bound of regulated power generations, and the upper bound of regulated power generations, etc.

[0074] The entity attributes in the operation scenarios of the Three Gorges Project include the name of the water conservancy project, the drawdown period, the flood season, the impoundment period, a certain month, a certain ten-day period, water level values, discharge values, power generation values, and condition values, etc. The relationship attributes include the characteristics of the water conservancy project, the operation stage, the operation cycle, the boundaries and constraints, and the operation conditions, etc.

[0075] (4) According to the characteristics of the Three Gorges Reservoir water conservancy project and the characteristics of different operation stages and scenarios, graph database software such as Neo4j and GraphDB is used to store and visualize these triple data, and the construction of the sub-graphs of the water conservancy project characteristics, operation stages, and operation scenarios is completed. A partial graph of the sub-graph of the Three Gorges Reservoir operation stage is shown as Figure 4 shown.

[0076] (5) Starting from the actual application requirements and the sub-graph structure, rules for knowledge fusion are defined to integrate the sub-graphs into the knowledge graph of the operation boundaries and constraints of the reservoir group, and the intelligent construction of the operation boundary and constraint library is completed.

[0077] It is easy for those skilled in the art to understand that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A knowledge graph construction method based on reservoir group scheduling boundaries and constraint libraries, characterized in that: The method comprises the following steps: S1 analyzes the unstructured data of reservoir operation regulations and the historical operation data of reservoirs to obtain the operation boundaries and constraints of different water conservancy project characteristics, reservoir operation control parameters in different operation stages and different scenarios; S2 uses the obtained different water conservancy project characteristics, the water level of the reservoir in different scheduling stages and different scenarios, the dispatching boundaries and constraints of the outflow flow and output to construct the corresponding sub-graphs, including the water conservancy project characteristic sub-graph, the scheduling stage sub-graph and the scheduling scenario sub-graph, each of which includes multiple entity nodes, each of which contains the entity attributes and entity content of the entity node, and the entity nodes are connected through the relationship attributes between the entity nodes; S3 merges the same entity nodes in the water conservancy project feature sub-graph, the scheduling stage sub-graph and the scheduling scenario sub-graph to achieve the fusion of each sub-graph, thereby obtaining the required knowledge graph.

2. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 1, characterized in that: The analysis of the reservoir dispatching regulations adopts a large language model, the input of the large language model is: prompt sentence + reservoir dispatching regulations text, the prompt sentence adopts the structure of the CO-STAR framework, and the output is the water conservancy project characteristics.

3. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 1 or 2, characterized in that: The historical dispatching and operating data of the reservoir is analyzed by using a sliding window statistical analysis method.

4. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 1, characterized in that: When constructing a sub-graph, the different water conservancy project characteristics, the water level of the reservoir in different scheduling stages and different scenarios, the scheduling boundaries and constraints of the outflow flow and output are first converted into ternaries. The ternaries include entity node A, entity node B, and the relationship attributes between entity nodes A and B. The entity content in the entity nodes A and B is displayed on the entity nodes of the sub-graph, and the entity attributes are not displayed.

5. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 4, characterized in that: The entity attributes in the water conservancy project characteristic sub-map include the name of the water conservancy project, the grade value of the water conservancy project, the function of the water conservancy project, the verification flood level value, the design flood level value, the flood control limit water level value, the low water level value during the dry season, the installed capacity value, the unit type, and the unit discharge capacity value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, characteristic water level, and generator unit characteristics.

6. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 4, characterized in that: The entity attributes in the scheduling stage sub-graph include the name of the water conservancy project, the drawdown period, the flood season, the water storage period, the water level value, the outflow flow value and the output value; the entity content is the name or value corresponding to each entity attribute; the relationship attributes include the characteristics of the water conservancy project, the scheduling stage, the lower limit of the historical water level, the upper limit of the historical water level, the lower limit of the historical flow, the upper limit of the historical flow, the lower limit of the historical output, the upper limit of the historical output, the lower limit of the regulation water level, the upper limit of the regulation water level, the lower limit of the regulation flow, the upper limit of the regulation flow, the lower limit of the regulation output and the upper limit of the regulation output.

7. A method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as claimed in claim 4, characterized in that: The entity attributes in the scheduling scenario sub-graph include the name of the water conservancy project, drawdown period, flood season, water storage period, a certain month, a certain ten-day period, water level value, outflow value and output value, and condition value; the entity content includes the name or value corresponding to each entity attribute; the relationship attributes include water conservancy project characteristics, scheduling stage, scheduling cycle, boundaries and constraints, and scheduling conditions.

8. The method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library according to claim 1, characterized in that: The operation control parameters include reservoir water level, reservoir outflow and reservoir output.

9. A system for constructing a knowledge graph using the knowledge graph construction method based on reservoir group scheduling boundaries and constraint libraries as described in any one of claims 1 to 8, characterized in that: The system includes a data acquisition module, a sub-graph construction module and a knowledge graph fusion module, among which: The data acquisition module is used to obtain different water conservancy project characteristics, the scheduling boundaries and constraints of the operation control parameters of the reservoir in different scheduling stages and different scenarios, and convert the acquired data into the corresponding ternary bodies; The knowledge graph construction module is used to construct a water conservancy project feature sub-graph, a scheduling stage sub-graph and a scheduling scenario sub-graph; The knowledge graph fusion module is used to fuse the sub-graphs to form the final knowledge graph.

10. A system for constructing a knowledge graph based on a knowledge graph construction method for a reservoir group scheduling boundary and constraint library, characterized in that: The system includes an executor, which is used to execute the method for constructing a knowledge graph based on a reservoir group scheduling boundary and constraint library as described in any one of claims 1-8.

Images