Cascade pump station group optimization operation scheme recommendation method and system based on knowledge graph

Abstract

The invention discloses a cascade pump station group optimization operation scheme recommendation method and system based on a knowledge graph, and the method comprises the steps: constructing a cascade pump station group historical operation case library; constructing a pump station group operation rule base; performing data structuring processing on unstructured data in the related data information to construct a mode layer of a cascade pump station group operation knowledge graph; based on the mode layer, performing data structuring processing on the cascade pump station group historical operation case library to construct a data layer of a cascade pump station group operation knowledge graph; and carrying out data storage on the extracted knowledge in a triple form based on the data layer, constructing a cascade pump station group operation knowledge graph, and carrying out search and operation scheme similarity matching on the cascade pump station group operation knowledge graph based on a pump station group operation rule base so as to recommend and obtain an optimization scheme of cascade pump station group operation according to a matching result. According to the method, the optimization scheme of cascade pump station group operation can be quickly recommended through knowledge graph searching and operation scheme similarity matching.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of optimal operation of cascade pumping station group, and particularly relates to a method and system for recommending optimal operation scheme of cascade pumping station group based on knowledge graph. BACKGROUND

[0002] Optimal operation of cascade pumping station group is one of the important measures to ensure the safety of water supply system, improve the efficiency and service level of water conservancy projects. Optimal operation of cascade pumping station group is beneficial to improve the efficiency of pumping station water supply, reduce energy consumption and cost, improve system stability and reliability, and realize intelligent management, improve the overall operation effect and service level of water supply system and promote the sustainable use of water resources. Knowledge graph is a new research method that integrates statistics, information science, semantic analysis, intelligent recognition and other technologies and theories. Knowledge graph can well serve the system monitoring and data analysis, establishment of optimization model, intelligent control and scheduling, fault detection and maintenance, risk warning and response in the recommendation of optimal operation scheme of pumping station group.

[0003] Currently, the recommendation method of optimal operation scheme of cascade pumping station group usually includes data collection and preprocessing, establishment of mathematical model, definition of optimization target, selection of appropriate optimization algorithm, optimization solving and evaluation and verification. First, through real-time monitoring data collection and preprocessing, the data quality and accuracy are ensured. Secondly, according to the characteristics and actual operation of pumping station group, appropriate mathematical model is established to describe the relationship and influence between pumping stations. Then the optimization target is defined, and appropriate optimization algorithm is selected, such as genetic algorithm, particle swarm optimization algorithm, etc., to search for the optimal solution. Subsequently, the optimization solving process is carried out, and the optimal or near-optimal solution that meets the optimization target is found through iterative calculation and parameter adjustment. Finally, the optimization results are evaluated and verified, and the feasibility and effectiveness of the recommended scheme are ensured through simulation test, comparative analysis and field verification.

[0004] Traditional methods usually make decisions based on expert experience and rules. The limitation of this method is that it cannot fully consider the complex system dynamics and nonlinear relationship, and cannot capture the potential optimization potential, resulting in inaccurate and inefficient recommended schemes. Traditional methods often focus on the optimization of a single target, such as economy or energy consumption. However, in practical applications, the optimal operation of cascade pumping station group involves the trade-off of multiple targets, such as economy, environmental impact, equipment life, etc. Traditional methods cannot provide comprehensive multi-objective optimization schemes, and often need to rely on manual adjustment and trial and error in the process of recommending schemes, by constantly trying and adjusting parameters to find the optimal solution. This method is time-consuming and labor-intensive, and may not find the global optimal solution. SUMMARY

[0005] The present application aims to at least partially solve one of the technical problems in the related art.

[0006] To this end, the application provides a knowledge graph-based cascade pump station group optimal operation scheme recommendation method, which is based on the principle of hydraulics and supported by the knowledge graph, and can effectively match the optimal operation scheme.

[0007] Another object of the application is to provide a knowledge graph-based cascade pump station group optimal operation scheme recommendation system.

[0008] To achieve the above object, the application provides a knowledge graph-based cascade pump station group optimal operation scheme recommendation method, which comprises the following steps:

[0009] Obtaining relevant data information of the cascade pump station group;

[0010] Constructing a historical operation case database of the cascade pump station group based on historical pump station group case data;

[0011] Constructing a pump station group operation rule database based on the historical operation case database of the cascade pump station group;

[0012] Structuring the unstructured data in the relevant data information to construct a mode layer of the cascade pump station group operation knowledge graph;

[0013] Structuring the historical operation case database of the cascade pump station group based on the mode layer to construct a data layer of the cascade pump station group operation knowledge graph;

[0014] Storing the extracted knowledge in the form of triples based on the data layer, constructing the cascade pump station group operation knowledge graph according to the data storage result, searching the cascade pump station group operation knowledge graph based on the pump station group operation rule database, and matching the operation scheme similarity to recommend the optimal operation scheme of the cascade pump station group according to the matching result.

[0015] The knowledge graph-based cascade pump station group optimal operation scheme recommendation method of the application can further have the following additional technical features:

[0016] In one embodiment of the application, constructing the historical operation case database of the cascade pump station group based on the historical pump station group case data comprises the following steps:

[0017] Obtaining historical operation data of the cascade pump station group based on the cascade pump station group system;

[0018] According to the characteristics and key indicators of the cascade pumping station group, a standard of historical case definition of the cascade pumping station group is formulated, historical operation data of the cascade pumping station group is input into a historical operation case library of the cascade pumping station group, and the input cases are analyzed and summarized to extract new operation cases of the cascade pumping station group;

[0019] The new operation cases of the cascade pumping station group are input to obtain a real-time historical operation case library of the cascade pumping station group.

[0020] In an embodiment of the present application, a pumping station group operation rule library is constructed based on the historical operation case library of the cascade pumping station group, comprising:

[0021] Based on the historical operation case library of the cascade pumping station group, pumping station group operation related information is obtained;

[0022] Based on the pumping station group operation related information, operation requirements and operation targets of the cascade pumping station group are determined, and constraint conditions and key performance indicators of the cascade pumping station group are analyzed to obtain data analysis results;

[0023] Based on the data analysis results, a relationship analysis result between data in the historical operation case library of the cascade pumping station group and the pumping station group operation rules is obtained, and a data-driven operation rule is extracted based on the relationship analysis result, so as to complete the construction of the pumping station group operation rule library.

[0024] In an embodiment of the present application, unstructured data in the related data information is subjected to data structuring processing to construct a schema layer of the cascade pumping station group operation knowledge graph, comprising:

[0025] A knowledge graph ontology model is constructed;

[0026] Based on the related data information of the cascade pumping station group, a cascade pumping station group information model is constructed;

[0027] By analyzing the knowledge structure in the cascade pumping station group information model, the knowledge graph ontology model is decomposed into a plurality of sub-concept models, multi-source information in the operation process is integrated into the sub-concept models, and concept attributes are described using an ontology language to construct the schema layer of the cascade pumping station group operation knowledge graph.

[0028] In an embodiment of the present application, the historical operation case library of the cascade pumping station group is subjected to data structuring processing to construct a data layer of the cascade pumping station group operation knowledge graph, comprising:

[0029] Based on the historical operation case library of the cascade pumping station group, cascade pumping station group operation related data is obtained;

[0030] Knowledge extraction is performed on the preprocessed cascade pumping station group operation related data to integrate and process data from different data sources to obtain data extraction results;

[0031] According to the field knowledge and requirements of the cascade pump station group operation, the data extraction result is modeled and designed to determine various data entities, and attribute definition results of entity attributes and relationship attributes are obtained;

[0032] According to the knowledge extraction data source established based on the pump station group typical operation characteristics extracted from the historical pump station group operation case library and the pump station group operation information element, and

[0033] Based on the semantic relationship of the mode layer and the attribute definition result, and by using information extraction, knowledge extraction, knowledge reasoning and storage mapping methods, a data layer of the cascade pump station group operation knowledge graph is constructed.

[0034] In an embodiment of the present application, the cascade pump station group operation knowledge graph is searched and the operation scheme similarity matching is matched based on the pump station group operation rule library, so as to recommend an optimized operation scheme of the cascade pump station group according to the matching result, including:

[0035] Based on the cascade pump station group operation knowledge graph, an initial condition of the cascade pump station group operation is obtained;

[0036] Based on the cascade pump station group operation rule library, a pump station group subgraph conforming to the initial condition is searched in the cascade pump station group operation knowledge graph, and the weight of the pump station group entity in the pump station group subgraph is calculated based on a similarity algorithm, so as to calculate the subgraph similarity according to the entity weight;

[0037] Based on the subgraph similarity, the cascade pump station group operation knowledge graph is searched and the operation scheme similarity matching is matched, and the scheme entity corresponding to each subgraph is weighted according to the similarity matching result, so as to recommend an optimized operation scheme of the cascade pump station group.

[0038] To achieve the above purpose, another aspect of the present application provides a cascade pump station group optimized operation scheme recommendation system based on a knowledge graph, including:

[0039] A historical information acquisition module is configured to acquire relevant data information of the cascade pump station group;

[0040] An operation case library construction module is configured to construct a historical operation case library of the cascade pump station group based on historical pump station group case data;

[0041] An operation rule library construction module is configured to construct a pump station group operation rule library based on the historical operation case library of the cascade pump station group;

[0042] A mode layer construction module is configured to perform data structuralization processing on unstructured data in the relevant data information to construct a mode layer of the cascade pump station group operation knowledge graph;

[0043] a data layer construction module configured to perform data structural processing on the historical operation case library of the cascade pump station group based on the schema layer to construct a data layer of the operation knowledge graph of the cascade pump station group;

[0044] an optimization scheme recommendation module configured to store the extracted knowledge in a form of triplets based on the data layer, to construct the operation knowledge graph of the cascade pump station group according to a data storage result, to search the operation knowledge graph of the cascade pump station group based on the operation rule library of the pump station group and to perform similarity matching on operation schemes, and to recommend an optimization scheme for the operation of the cascade pump station group according to a matching result.

[0045] The method and system for recommending an optimization operation scheme of a cascade pump station group based on a knowledge graph according to the embodiments of the present application can quickly and effectively recommend an optimization operation scheme of the cascade pump station group by searching the knowledge graph and performing similarity matching on operation schemes, through inputting actual conditions and operation condition related data of the cascade pump station group.

[0046] Additional aspects and advantages of the present application will be made apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0047] The above and / or additional aspects and advantages of the present application will become apparent and be more readily understood from the following description, taken in conjunction with the accompanying drawings, in which:

[0048] Figure 1 is a flowchart of a method for recommending an optimization operation scheme of a cascade pump station group based on a knowledge graph according to an embodiment of the present application;

[0049] Figure 2 is an architecture diagram of a method for recommending an optimization operation scheme of a cascade pump station group based on a knowledge graph according to an embodiment of the present application;

[0050] Figure 3 is a structure schematic diagram of a system for recommending an optimization operation scheme of a cascade pump station group based on a knowledge graph according to an embodiment of the present application. DETAILED DESCRIPTION

[0051] It should be noted that the embodiments and features of the present application can be combined with each other without conflict, and the present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0052] In order for those skilled in the art to better understand the technical scheme of the present application, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0053] The knowledge graph-based cascade pumping station group optimal operation scheme recommendation method and system proposed according to the embodiments of the present application are described below with reference to the drawings.

[0054] Figure 2 The architecture diagram of the knowledge graph-based cascade pumping station group optimal operation scheme recommendation method of the embodiments of the present application is shown in the figure. Based on information demand analysis, data collection and sorting, and data modeling, the subordinate pumping stations, total water pumping volume, number of pumping stations, pumping station types, pumping station levels, water pump units, subordinate buildings, and other related data information of the cascade pumping station group are obtained, and a cascade pumping station group information model is constructed. Based on the existing pumping station group case data or the actual pumping station group operation process case data, the related case data is analyzed and summarized, and a cascade pumping station group historical operation case library is constructed. Based on the constructed cascade pumping station group historical operation case library, the existing pumping station group operation rules and operation processes are sorted through multiple information sources such as pumping station operation and maintenance personnel and technical experts, a pumping station group operation rule library is formed, and a cascade pumping station group operation knowledge and experience module is constructed. Based on the constructed cascade pumping station group operation knowledge and experience module, the pumping station group operation parameter data, experience case data, equipment information data, and operation and maintenance record data therein are analyzed through concept analysis, ontology modeling, and semantic rule extension, and the construction of the mode layer of the cascade pumping station group operation knowledge graph is completed. Based on the constructed cascade pumping station group historical operation case library, the construction of the data layer of the cascade pumping station group operation knowledge graph is completed by using information extraction, knowledge extraction, knowledge reasoning, and storage mapping. Based on the constructed cascade pumping station group operation knowledge graph, the knowledge graph is searched and the operation scheme similarity is matched in combination with the pumping station group operation rule library, so as to recommend the optimal operation scheme of the cascade pumping station group, as shown in the figure. Figure 2

[0055] Figure 1 The flowchart of the knowledge graph-based cascade pumping station group optimal operation scheme recommendation method of the embodiments of the present application is shown in the figure.

[0056] As shown in the figure, the method includes but is not limited to the following steps: Figure 1

[0057] S1, obtaining the related data information of the cascade pumping station group.

[0058] ​​Specifically, based on information demand analysis, data collection and organization, data modeling, the subordinate pump stations of the cascade pump station group, the total amount of water pumping, the number of pump stations, the type of pump stations, the level of pump stations, water pump units, and subordinate buildings are obtained.

[0059] Further, the cascade pump station group information model can help to structure the data related to the cascade pump station group, including pump station attributes, water conservancy engineering parameters, and running status, so as to facilitate the management, analysis and application of the system.

[0060] Further, through the cascade pump station group information model, the correlation between different data can be established, providing a comprehensive data view for the system, which is helpful for comprehensive analysis of the operation status and efficiency of the pump station group.

[0061] S2, based on historical pump station group case data, a historical operation case library of the cascade pump station group is constructed.

[0062] Specifically, from the existing cascade pump station group operation and maintenance system, equipment monitoring system or related files, the historical operation data of the cascade pump station group in the past (including the operation record of each pump station, fault condition, optimization record, energy consumption data, etc.) is collected. The collected initial data of the cascade pump station group operation is sorted and archived, classified and archived according to the pump station and time sequence, to ensure the completeness and accuracy of the data, and a database or a special data management software is used to organize and store the data.

[0063] Specifically, according to the characteristics and key indicators of the cascade pump station group, different types of optimization cases are defined (including optimization target, scheme design, implementation process, optimization result, etc.), a set of standards for defining historical cases of the cascade pump station group is formulated, and the initial data of the cascade pump station group operation is entered into the pump station group case library one by one, providing detailed description for each case. By observing the commonness, regularity and trend in the historical cases of the pump station group, data analysis tools and statistical methods are used to assist analysis, the entered cases are analyzed and summarized, and valuable operation experience and lessons of the cascade pump station group (operation and maintenance best practice experience, fault handling experience, optimization improvement experience, risk prevention lessons) are extracted.

[0064] Specifically, as time goes by and new operation cases appear, new cases are added to the case library in time and updated, the case library is reviewed and maintained regularly to ensure that the data and information in it are up-to-date. Accordingly, according to the business needs and specific circumstances, the actual application of the case library is adjusted and optimized to obtain a more perfect historical operation case library of the cascade pump station group.

[0065] S3, based on the historical operation case library of the cascade pump station group, a pump station group operation rule library is constructed.

[0066] Specifically, based on the historical operation case library of the cascade pumping station group constructed by S2, knowledge and experience about the operation of the cascade pumping station group (including various operation processes, optimized operation schemes, performance optimization techniques, safety management measures, etc.) are collected from multiple information sources such as pumping station operation and maintenance personnel, engineers, technical experts, etc.; the collected knowledge and experience are sorted and classified, and expressed in the form of text description, diagram, flowchart, operation manual, etc. to clearly express the knowledge and experience, and the sorted knowledge and experience are expressed and documented to ensure clear document structure and accurate and complete information.

[0067] Specifically, based on the collected information related to the operation of the pumping station group (including the structure of the pumping station group, equipment parameters, characteristics of the water supply system, pipe network conditions, etc.), the operation requirements of the cascade pumping station group are determined according to factors such as water supply, pressure requirements, flood control requirements, etc., the operation targets are determined according to water supply, pressure requirements, flood control requirements, etc., and the constraint conditions and key performance indicators are analyzed based on the clear operation targets. Correspondingly, the existing operation rules and operation processes of the pumping station group (including industry standards, technical specifications, operation manuals, etc.) are sorted and evaluated, and the obsolete, repetitive or inapplicable rules are removed. At the same time, experience and professional knowledge are obtained from experts, engineers, operation and maintenance personnel, etc. in related fields, industry best practices, common problems and solutions are understood, and their experience is used as part of the rule library. Through analysis of the data in the historical operation case library of the cascade pumping station group, a mathematical model is established, and the relationship between the data and the operation rules of the pumping station group is explored. Based on the data analysis results, data-driven operation rules are extracted, and the construction of the operation rule library of the cascade pumping station group is completed. Finally, the operation rule library of the cascade pumping station group is documented and managed (including writing detailed rule explanations and usage guidelines), a version management and update mechanism for the rule library is established, and the rule library is continuously updated and improved over time and with the accumulation of operation experience. The specific construction process can be adjusted and customized according to the actual situation of the cascade pumping station group, and close communication and cooperation with operation and maintenance personnel, experts and relevant departments are required during the construction process to ensure the practicality and applicability of the rule library.

[0068] Specifically, a storage, management and retrieval mechanism for the knowledge base is established based on the sorted knowledge and experience documents of the cascade pumping station group, and an electronic document management system, knowledge management software or online platform is used to build and manage the knowledge base, making it convenient for pumping station group operation and maintenance personnel to consult and use; at the same time, with the development of pumping station operation and technology, the knowledge and experience base is continuously updated and iterated, new knowledge and experience is incorporated into the base in a timely manner, ensuring that the pumping station group knowledge base keeps pace with the actual operation and maintenance work of the pumping station group, and ensuring the accuracy and completeness of the knowledge and experience, thereby establishing the cascade pumping station group operation knowledge and experience module.

[0069] S4, data structuring processing is performed on unstructured data in the related data information to construct a mode layer of the cascade pump station group operation knowledge graph.

[0070] Specifically, the knowledge graph mode layer is a conceptual model of knowledge graph construction, which can standardize and constrain data. In order to enhance the plasticity of the knowledge graph and standardize the knowledge in the field, generally, modeling work of the knowledge in the field is needed before constructing the knowledge graph. Generally, knowledge modeling is divided into top-down and bottom-up. In the process of ontology concept knowledge modeling, the scope of the knowledge graph involved needs to be defined first, for example, the field of water conservancy engineering. Then the specific concepts that need to be modeled are determined, such as pump stations, reservoirs, pipe networks, etc. Each concept class (class) is defined using ontology language (such as OWL), for example, pump stations, reservoirs, etc. Secondly, the relations between classes (concepts) are described, such as the water supply relation between pump stations and reservoirs, the connection relation between pump stations and pipe networks, etc. The characteristics and state of the concept are described by defining the attributes of the concept class, including instance-level attributes and class-level attributes. Some axioms of the ontology are formulated to specify the logical relations between concepts, such as defining that a pump station must be connected to a pipe network to operate normally. Finally, specific instances are added to the ontology, such as a specific pump station entity, a specific reservoir entity, etc. Relations are used to describe the relations between classes (concepts), such as part-of, kind-of, etc., which are structured defined as follows:

[0071] O=[C, P C ,R, P R ,A]

[0072] Where C represents all concepts in the described knowledge, PC represents all attribute values of the concepts in the described knowledge, R represents all relations of the concepts in the described knowledge, P R represents all attributes in the relations, and A represents the axiom set.

[0073] Specifically, the operation knowledge and experience module of the cascade pumping station group constructed based on S3 constructs a mode layer of the cascade pumping station group through existing data information of the cascade pumping station group. First, an information model of the cascade pumping station group is constructed, which is composed of information such as subordinate pumping stations, total water lifting amount, and number of pumping stations. The pumping station information can be further divided into information such as type, level, water pump unit, and subordinate buildings. The water pump unit can be further divided into information such as number of units, rated power of units, blade setting angle of each unit at each time period, rotating speed and operating lift of units, and operating rules. Among them, by analyzing the knowledge structure in the information model, the knowledge graph ontology model is decomposed into multiple sub-concept models, and the multi-source information in the operation process is gradually integrated into these concepts. Then, the concept attributes are described using the ontology language OWL, and finally an ontology model with rich semantic relationships and clear hierarchical structure is constructed to obtain the mode layer of the operation knowledge graph of the cascade pumping station group, which is constructed in combination with concepts such as total water lifting head of the cascade pumping station group and cascade number, and can be represented as:

[0074]

[0075] Among them, E TJ is the ontology model of the cascade pumping station group, E XS is the ontology model of the subordinate pumping station of the cascade pumping station group, TJ is the number of stages contained by the cascade pumping station group, E h is the concept ontology of the total water lifting head of the cascade pumping station group, E n is the concept ontology of the cascade number, and E else is other concept ontologies related to the cascade pumping station group.

[0076] S5, based on the mode layer, data structural processing is performed on the historical operation case library of the cascade pumping station group to construct the data layer of the operation knowledge graph of the cascade pumping station group.

[0077] Specifically, based on the historical operation case library of the cascade pumping station group, various data sources related to the operation of the cascade pumping station group (including operation records, equipment parameters, sensor data, maintenance logs, optimization records, etc.) are collected through installation of a monitoring system, review of operation and maintenance records of the cascade pumping station group, communication and exchange with operation and maintenance personnel and experts of the pumping station group, etc. The collected operation data of the cascade pumping station group are cleaned and preprocessed (including removal of duplicate, missing, or incorrect data), and then knowledge extraction is performed to integrate and fuse data from different data sources and establish the association relationship between data, which can be realized through shared identifiers, timestamps, etc. to ensure the consistency and integrity of the data.

[0078] Among them, knowledge extraction refers to automatically extracting data of different sources and structures in cascade pumping station group operation data, converting non-structured data with comprehensive attributes into more structured data, and storing them in the cascade pumping station group knowledge graph. It is of great significance to build a comprehensive relationship of the pumping station group knowledge graph. The data sources of knowledge extraction mainly include three parts of structured data, semi-structured data and unstructured data in the cascade pumping station group operation data. The first part of the pumping station group structured data refers to the pumping station group data expressed in two-dimensional structure table, which follows the specification of data format and length, is stored in a relational database, and is managed, such as link data, database data. The second part of the pumping station group semi-structured data is a form of structured data, which contains relevant tags to isolate semantic units and layer records and fields. The third part of the pumping station group unstructured data has the characteristics of irregularity and incompleteness, and has no specific model, such as all formats of office documents, texts, pictures, HTML, various reports, images and video information, etc. Correspondingly, for different structures of data sources in the cascade pumping station group, the entities, attributes and relationship data required to build the cascade pumping station group operation knowledge graph are extracted.

[0079] Specifically, according to the domain knowledge and requirements of the cascade pumping station group operation, the extracted entities, attributes and relationships of the cascade pumping station group operation knowledge graph are modeled and structured, the relationships between various data entities (such as pumping stations, equipment, optimization records, etc.) are determined, and the entity attributes and relationship attributes are defined. Correspondingly, appropriate data storage and management methods are selected, such as relational databases, graph databases or other storage systems, and according to the data modeling results, the corresponding tables, collections or graph structures are created, and data import and indexing operations are performed. Quality control is performed on the stored data, including data integrity, consistency and accuracy checking, which can use data verification, anomaly detection and other methods to identify and handle data quality problems.

[0080] Specifically, the data layer construction of the cascade pumping station group operation knowledge graph includes two parts of feature-operation model construction and knowledge extraction and fusion. The feature-operation information model extracts the typical operation features of the pumping station group and the pumping station group operation information from the historical pumping station group operation case library to establish the data source of knowledge extraction. Knowledge extraction and fusion is to construct the data layer of the knowledge graph by using information extraction, knowledge extraction, knowledge reasoning and storage mapping according to the semantic relationship provided by the mode layer. Finally, all the knowledge extracted from the pumping station group operation cases is stored in the form of triples to form the pumping station group operation knowledge graph.

[0081] S6, based on the data layer, storing the extracted knowledge in the form of triples, constructing a cascade pumping station group operation knowledge graph according to the data storage result, searching the cascade pumping station group operation knowledge graph based on the pumping station group operation rule library, and matching the operation scheme similarity, to recommend an optimized scheme for the cascade pumping station group operation according to the matching result.

[0082] Specifically, based on the constructed cascade pumping station group operation knowledge graph, the initial conditions of the cascade pumping station group operation are obtained, including the pumping station layout (including the position, number, arrangement, relative position of each pumping station, and the relationship between them, etc.), the pumping station equipment (including the type, specification, and performance parameters of the pump, valve, and pipeline equipment), the working mode (including the flow demand, pressure demand, and other operation mode requirements), the configuration strategy (including the control strategy and scheduling strategy of the pumping station group), the dynamic response requirement (including the response speed requirement of the pumping station group to flow changes and pressure changes), the power supply (including the power supply situation of the pumping station group, the voltage, frequency, and other elements of the power supply), the control system (including the control system hardware and software of the pumping station group, the type, function, and performance requirement of the control system), the environmental conditions (including the climate and water quality of the environment where the pumping station group is located), which play an important role in the design and operation process and have an important influence on the stability and reliability of the pumping station group. Therefore, these initial conditions need to be carefully considered and reasonably determined in practical applications to form a cascade pumping station group subgraph.

[0083] Specifically, combined with the cascade pumping station group operation rule library, such as the pumping station water volume optimization rule, the water level-lift optimization principle, the joint operation economy principle of the pumping station group, and the inter-stage water conveyance river water level requirement, the pumping station group subgraph with similar initial operation conditions is searched in the knowledge graph composed of the pumping station group historical cases; the weight of the pumping station group entity in each pumping station group subgraph is calculated based on the similarity, and the subgraph similarity is calculated according to the entity weight.

[0084] Among them, the knowledge graph subgraph similarity calculation is a method for measuring the similarity between two knowledge graph subgraphs. First, select the constructed cascade pump station group operation knowledge graph from the two knowledge graphs as the reference graph, and then select one or more nodes as the root node of the subgraph. Each node and the corresponding relationship is represented as a feature vector, and a feature representation method based on attributes, topological structure, etc. is used to ensure that the feature vectors of nodes and relationships can accurately express their attributes and associated information. A suitable similarity measurement method is used to calculate the similarity between the two subgraphs, which includes: representing the subgraph as a combination of graph structure and feature vector, and then calculating the similarity between the feature vectors of the two subgraphs to measure the similarity or according to the similarity measurement result, the similarity of the two subgraphs can be evaluated to obtain a similarity score or similarity index. According to the specific requirements, the threshold value can be set to judge whether the two subgraphs are similar or the degree of similarity. At the same time in practical application, according to the specific situation, the appropriate method is selected, and the verification and adjustment are carried out to obtain more accurate similarity calculation result.

[0085] Specifically, based on the constructed cascade pump station group operation knowledge graph, various cascade pump station group operation schemes under various conditions are provided, and based on the actual situation and operation conditions of the pump station group, the knowledge graph is searched and the operation scheme similarity is matched, the subgraph similarity is calculated, and the scheme entity corresponding to each subgraph is weighted according to the triple path query of the knowledge graph, so as to recommend the optimization scheme of the cascade pump station group operation.

[0086] The method for recommending the optimization scheme of the cascade pump station group operation based on the knowledge graph according to the embodiment of the application improves the intelligent recommendation method of the optimization scheme of the pump station group with the cascade pump station group as the core, adds the pump station group information module based on data acquisition, transmission, processing and display, adds the pump station group operation knowledge and experience module based on knowledge acquisition, knowledge representation, knowledge reasoning, individual adaptation and experience accumulation and updating, and adds the pump station group historical operation case library module based on the acquisition, cleaning, storage and query analysis of case data. By inputting the actual situation and operation condition related data of the cascade pump station group, the application can quickly and effectively recommend the optimization scheme of the cascade pump station group operation by searching the knowledge graph and matching the operation scheme similarity.

[0087] In order to realize the above-mentioned embodiment, as Figure 3 shown, the embodiment also provides a system 10 for recommending the optimization scheme of the cascade pump station group operation based on the knowledge graph, which comprises a historical information acquisition module 100, an operation case library construction module 200, an operation rule library construction module 300, a mode layer construction module 400, a data layer construction module 500 and an optimization scheme recommendation module 600.

[0088] The historical information acquisition module 100 is configured to acquire relevant data information of the cascade pump station group.

[0089] The operation case library construction module 200 is configured to construct a historical operation case library of the cascade pump station group based on historical pump station group case data.

[0090] The operation rule library construction module 300 is configured to construct a pump station group operation rule library based on the historical operation case library of the cascade pump station group.

[0091] The schema layer construction module 400 is configured to perform data structuring processing on unstructured data in the relevant data information to construct a schema layer of the cascade pump station group operation knowledge graph.

[0092] The data layer construction module 500 is configured to perform data structuring processing on the historical operation case library of the cascade pump station group based on the schema layer to construct a data layer of the cascade pump station group operation knowledge graph.

[0093] The optimization scheme recommendation module 600 is configured to store the extracted knowledge in the form of triples based on the data layer, to construct a cascade pump station group operation knowledge graph according to a data storage result, and to search the cascade pump station group operation knowledge graph and perform similarity matching on operation schemes based on the pump station group operation rule library, so as to recommend an optimization scheme for operation of the cascade pump station group according to a matching result.

[0094] Further, the operation case library construction module 200 is further configured to:

[0095] acquire historical operation data of the cascade pump station group based on a cascade pump station group system;

[0096] formulate a standard for defining historical cases of the cascade pump station group according to characteristics and key indicators of the cascade pump station group, to input the historical operation data of the cascade pump station group into the historical operation case library of the cascade pump station group, and analyze and summarize the input cases to extract new operation cases of the cascade pump station group;

[0097] input the new operation cases of the cascade pump station group to obtain a real-time historical operation case library of the cascade pump station group.

[0098] Further, the operation rule library construction module 300 is further configured to:

[0099] acquire pump station group operation related information based on the historical operation case library of the cascade pump station group;

[0100] determine operation requirements and operation targets of the cascade pump station group based on the pump station group operation related information, and analyze constraint conditions and key performance indicators of the cascade pump station group to obtain data analysis results;

[0101] The relationship analysis result is obtained based on data analysis result analysis of data in the historical operation case library of the cascade pump station group and the operation rule of the pump station group, and the data-driven operation rule is extracted based on the relationship analysis result, so as to complete construction of the operation rule library of the cascade pump station group.

[0102] Further, the mode layer construction module 400 is further used for:

[0103] constructing a knowledge graph ontology model;

[0104] constructing a cascade pump station group information model based on related data information of the cascade pump station group;

[0105] The knowledge graph ontology model is decomposed into a plurality of sub-concept models by analyzing the knowledge structure in the cascade pump station group information model, the multi-source information in the operation process is integrated into the sub-concept models, and the concept attributes are described using an ontology language to construct a mode layer of the cascade pump station group operation knowledge graph.

[0106] The knowledge graph-based cascade pump station group optimal operation scheme recommendation system according to the embodiment of the present application perfects the intelligent recommendation method of the pump station group optimal operation scheme with the cascade pump station group as the core, adds the pump station group information module based on data acquisition, transmission, processing and display, adds the pump station group operation knowledge and experience module based on knowledge acquisition, knowledge representation, knowledge reasoning, individualized adaptation and experience accumulation and updating, and adds the pump station group historical operation case library module based on case data acquisition, cleaning, storage and query analysis. By inputting the actual situation and operation condition related data of the cascade pump station group, the present application can quickly and effectively recommend the optimal operation scheme of the cascade pump station group through knowledge graph search and operation scheme similarity matching.

[0107] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0108] Furthermore, the terms "first", "second", "third", "fourth", "fifth" and "sixth" are used herein for descriptive purposes only and are not to be construed as indicating or implying relative importance or a significant nature of so described technical features. It is to be understood that a technical feature described with the "first", "second", "third", "fourth", "fifth" or "sixth" can implicitly or explicitly include at least one of the technical features described with the "first", "second", "third", "fourth", "fifth" or "sixth". In the description of the present application, the meaning of "a plurality" is at least two, for example, two, three, etc., unless otherwise specifically defined.

Claims

1. A knowledge graph-based step-by-step pump station group optimal operation scheme recommendation method, characterized in that, The method comprises the following steps: obtaining relevant data information of the cascade pumping station group; constructing a historical operation case library of the cascade pumping station group based on historical case data of the pumping station group; constructing a pumping station group operation rule library based on the historical operation case library of the cascade pumping station group; performing data structuring processing on unstructured data in the relevant data information to construct a mode layer of a cascade pumping station group operation knowledge graph; performing data structuring processing on the historical operation case library of the cascade pumping station group based on the mode layer to construct a data layer of the cascade pumping station group operation knowledge graph; performing data storage on the extracted knowledge in the form of triples based on the data layer, constructing a cascade pumping station group operation knowledge graph according to the data storage result, and performing search and operation scheme similarity matching on the cascade pumping station group operation knowledge graph based on the pumping station group operation rule library to obtain an optimized operation scheme of the cascade pumping station group according to the matching result.

2. The method of claim 1, wherein, The construction of the historical operation case library of the cascade pumping station group based on historical case data of the pumping station group comprises: obtaining historical operation data of the cascade pumping station group based on the cascade pumping station group system; defining standards for the definition of historical cases of the cascade pumping station group according to the characteristics and key indicators of the cascade pumping station group, entering the historical operation data of the cascade pumping station group into the historical operation case library of the cascade pumping station group, and analyzing and summarizing the entered cases to extract new operation cases of the cascade pumping station group; performing data entry on the new operation cases of the cascade pumping station group to obtain a real-time historical operation case library of the cascade pumping station group.

3. The method of claim 1, wherein, The construction of the pumping station group operation rule library based on the historical operation case library of the cascade pumping station group comprises: obtaining pumping station group operation related information based on the historical operation case library of the cascade pumping station group; determining the operation requirements and operation targets of the cascade pumping station group based on the pumping station group operation related information, and analyzing the constraint conditions and key performance indicators of the cascade pumping station group to obtain data analysis results; analyzing the relationship between the data in the historical operation case library of the cascade pumping station group and the pumping station group operation rules based on the data analysis results to obtain relationship analysis results, and extracting data-driven operation rules based on the relationship analysis results to complete the construction of the pumping station group operation rule library.

4. The method of claim 3, wherein, The data structuring processing on the unstructured data in the relevant data information to construct the mode layer of the cascade pumping station group operation knowledge graph comprises: constructing a knowledge graph ontology model; constructing a cascade pumping station group information model based on the relevant data information of the cascade pumping station group; decomposing the knowledge graph ontology model into multiple sub-concept models by analyzing the knowledge structure in the cascade pumping station group information model, integrating multi-source information in the operation process into the sub-concept models, and describing the concept attributes using an ontology language to construct the mode layer of the cascade pumping station group operation knowledge graph.

5. The method of claim 4, wherein, The data structuring processing on the historical operation case library of the cascade pumping station group to construct the data layer of the cascade pumping station group operation knowledge graph comprises: obtaining cascade pumping station group operation related data based on the historical operation case library of the cascade pumping station group; performing knowledge extraction on the preprocessed cascade pumping station group operation related data to integrate and process data from different data sources to obtain data extraction results; According to the domain knowledge and requirements of the cascade pump station group operation, the data extraction result is modeled and designed to determine each data entity, and attribute definition results of entity attributes and relationship attributes are obtained; A data source for knowledge extraction is established according to the pump station group typical operation characteristics extracted from the historical pump station group operation case library and the pump station group operation information element; and Based on the semantic relationship of the schema layer and the attribute definition result, an information extraction, knowledge extraction, knowledge reasoning and storage mapping method are used to construct a data layer of the cascade pump station group operation knowledge graph.

6. The method of claim 5, wherein, Based on the pump station group operation rule library, the cascade pump station group operation knowledge graph is searched and the operation scheme similarity is matched, so as to recommend an optimized scheme for the cascade pump station group operation according to the matching result, including: Obtaining initial conditions for the cascade pump station group operation based on the cascade pump station group operation knowledge graph; Searching the cascade pump station group operation knowledge graph based on the initial conditions and the pump station group operation rule library, and calculating the weights of the pump station group entities in the pump station group subgraph based on a similarity algorithm, so as to calculate the subgraph similarity according to the entity weight; Searching the cascade pump station group operation knowledge graph based on the subgraph similarity and the operation scheme similarity matching, and weighting the scheme entities corresponding to each subgraph according to the similarity matching result to recommend an optimized scheme for the cascade pump station group operation.

7. A knowledge graph-based step-by-step pump station group optimal operation scheme recommendation system, characterized in that, It includes: A historical information acquisition module is configured to acquire relevant data information of the cascade pump station group; A running case library construction module is configured to construct a historical running case library of the cascade pump station group based on historical pump station group case data; A running rule library construction module is configured to construct a pump station group running rule library based on the historical running case library of the cascade pump station group; A schema layer construction module is configured to structure the unstructured data in the relevant data information to construct a schema layer of the cascade pump station group operation knowledge graph; A data layer construction module is configured to structure the historical running case library of the cascade pump station group based on the schema layer to construct a data layer of the cascade pump station group operation knowledge graph; An optimization scheme recommendation module is configured to store the extracted knowledge in the form of triples based on the data layer, to construct a cascade pump station group operation knowledge graph according to the data storage result, and to search the cascade pump station group operation knowledge graph based on the pump station group operation rule library and match the operation scheme similarity, so as to recommend an optimized scheme for the cascade pump station group operation according to the matching result.

8. The system of claim 7, wherein, The running case library construction module is further configured to: Acquire historical running data of the cascade pump station group based on the cascade pump station group system; Formulate standards for the definition of historical cases of the cascade pump station group according to the characteristics and key indicators of the cascade pump station group, so as to input the historical running data of the cascade pump station group into the historical running case library of the cascade pump station group, and analyze and summarize the input cases to extract new cascade pump station group operation cases; Input the new cascade pump station group operation cases to obtain a real-time historical running case library of the cascade pump station group.

9. The system of claim 8, wherein, The running rule library construction module is further configured to: Acquire pump station group operation related information based on the historical running case library of the cascade pump station group; determine operation requirements and operation targets of the cascade pump station group based on the information related to operation of the pump station group, and analyze constraint conditions and key performance indicators of the cascade pump station group to obtain data analysis results; analyze a relationship between data in a historical operation case library of the cascade pump station group and operation rules of the pump station group based on the data analysis results to obtain relationship analysis results, and extract data-driven operation rules based on the relationship analysis results to complete construction of an operation rule library of the cascade pump station group.

10. The system of claim 9, wherein, The mode layer construction module is further configured to: construct a knowledge graph ontology model; construct an information model of the cascade pump station group based on related data information of the cascade pump station group; decompose the knowledge graph ontology model into a plurality of sub-concept models by analyzing a knowledge structure in the information model of the cascade pump station group, integrate multi-source information in an operation process into the sub-concept models, and describe concept attributes using an ontology language to construct a mode layer of the knowledge graph of the operation of the cascade pump station group.

Images