An extensible method for constructing a digital twin of a water pump

Abstract

The present invention discloses an extensible method for constructing a digital twin of a water pump, specifically related to the field of digital twins, including S1: constructing a physical model, S2: collecting water pump performance data, S3: preprocessing the water pump performance data, S4: constructing a digital twin model of the water pump, S5: applying the digital twin model of the water pump, and S6: optimizing the digital twin of the water pump. By constructing the digital twin model of the water pump from multiple dimensions such as water pump flow data, water pump pressure data, and water pump operation data, and continuously improving the model according to the actual operation of the water pump, the extensibility of the digital twin of the water pump is realized. The digital twin model of the water pump can also perform safety monitoring on water resource allocation projects, judge different operating conditions of the water pump, and make different response results, further improving the safety of the project.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twin, and more specifically, to a method for constructing an extensible digital twin of a water pump. Background Art

[0002] As the core and key of the entire water resource allocation project, water pumps play a crucial role in water conservancy projects. In recent years, digital twin technology has received extensive attention and emphasis, and has been applied to varying degrees in various industries and fields. However, it has just started in the water conservancy industry, and its sustainable development still faces many key technologies that need to be solved urgently. How to combine digital twin technology with classical hydrodynamics theory to provide decision-making support for project operation management is an important issue that needs to be solved in the "New Era Ecological and Intelligent Water Conservancy Project".

[0003] The existing safety monitoring methods for water pumps in water conservancy projects generally include data collection and transmission steps, data processing and analysis steps, and monitoring stations and control centers. Among them, the data collection and transmission steps consist of data collectors and transmission devices, which are responsible for collecting data from each monitoring point and performing transmission and storage; the data processing and analysis steps include data processing software, data analysis algorithms, data models, and visualization tools, etc., to process, sort, and analyze the collected hydrological data; the monitoring stations of the monitoring stations and control centers are the installation points of various monitoring devices, providing real-time monitoring data, and the control center is the core of data collection, processing, analysis, and report generation, responsible for receiving and managing data from the monitoring stations, and providing corresponding hydrological information and services to relevant departments and users.

[0004] However, in actual use, there are still some disadvantages. For example, in the data collection and transmission steps, due to the water pump being in a complex and changeable natural environment, the uncertainty of extreme weather and climate increases the difficulty of monitoring and data inaccuracy; when collecting data, manual observation and simple sensor monitoring are often used, which cannot provide real-time and accurate data, and the limitations of sensor technology also limit the accuracy and scope of monitoring, making it impossible to detect some potential safety hazards in time; when analyzing data, due to different data collection departments, there are obstacles in data sharing and communication, resulting in difficulty in integrating and utilizing data. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a method for constructing an extensible digital twin of a water pump, through the following solutions to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] S1: Construct a physical model: Construct a physical model according to the water pump required for the water resource allocation project;

[0008] S2: Collect pump performance data: Based on the physical model constructed in S1, collect pump performance data for the water resource allocation project. The pump performance data includes pump flow data, pump pressure data, and pump operation data.

[0009] S3: Preprocess pump performance data: Preprocess the pump performance data collected in S2, including data cleaning, data correction, filtering, and feature extraction.

[0010] S4: Build a pump digital twin model: Based on the pump performance data and physical model preprocessed in S3, establish a pump performance intelligent analysis model, and build a pump digital twin model according to the pump performance intelligent analysis model.

[0011] S5: Apply the pump digital twin model: Based on the pump digital twin model constructed in S4, apply the pump digital twin model and perform visualization processing.

[0012] S6: Optimize the pump digital twin: Based on the pump performance data set, the pump performance intelligent analysis model, and the pump operation status, optimize the pump digital twin modularly.

[0013] Preferably, S2 includes a pump flow data collection unit, a pump pressure data collection unit, and a pump operation data collection unit. The pump flow data collection unit is used to collect pump flow data, including inlet flow, outlet flow, volumetric flow, and mass flow. The pump pressure data collection unit is used to collect pump pressure data, including inlet pressure, outlet pressure, and head. The pump operation data collection unit is used to collect pump operation data, including the historical operation time, historical maintenance time, pump active power, and shaft power of the pump.

[0014] Preferably, the preprocessing automatically screens and eliminates abnormal data in the pump performance data, and uses data cleaning technology to achieve data correction. Data cleaning refers to discovering and correcting identifiable errors in the data, including checking data consistency, handling invalid values and missing values. Filtering is to remove noise and interference in the pump performance data of the virtual body. Feature extraction is to extract information from the pump performance data collected in S2, including but not limited to the following features: Average value: Calculate the average historical maintenance time and average historical operation time over a period of time to reflect the operation status of the pump. Maximum and minimum values: Calculate the maximum and minimum outlet pressures and inlet pressures over a period of time to reflect the extreme working state of the pump. Percentage: Calculate the percentages of the pump, hydraulics, and volume over a period of time to reflect the efficiency of the pump.

[0015] Preferably, the digital twin model of the water pump is simulated based on the preprocessed water pump performance data, creating a virtual modeling corresponding to the physical world, establishing an intelligent analysis model for water pump performance, and establishing an intelligent analysis model for water pump performance based on the water pump knowledge graph. The intelligent analysis model for water pump performance includes, but is not limited to, a water pump flow digital model, a water pump pressure digital model, and a water pump operation digital model. The water pump flow digital model is used to retrieve the water pump flow data in the water pump performance dataset and analyze the water pump flow data to obtain a water pump flow stability evaluation coefficient. The specific analysis formula for the water pump flow stability evaluation coefficient is:

[0016]

[0017] where Y represents the water pump flow stability evaluation coefficient, Co represents the water pump outlet flow, Ci represents the water pump inlet flow, Cv represents the volumetric flow rate of the water pump, Cm represents the mass flow rate of the water pump, and Cv 预 represents the preset volume of liquid passing through the water pump outlet per unit time, and Cm 预 represents the preset mass of liquid passing through the water pump outlet per unit time; the water pump pressure digital model is used to retrieve the water pump pressure data in the water pump performance dataset and analyze the water pump pressure data to obtain a water pump pressure safety evaluation coefficient. The specific analysis formula for the water pump pressure safety evaluation coefficient is:

[0018]

[0019] where Z represents the water pump pressure safety evaluation coefficient, Lc represents the head of the water pump, and Lc 理 represents the theoretical head of the water pump, Po represents the outlet pressure of the water pump, Pi represents the inlet pressure of the water pump, and Po min represents the minimum outlet pressure, and Po max represents the maximum outlet pressure; the water pump operation digital model is used to retrieve the water pump operation data in the water pump performance dataset and analyze the water pump operation data to obtain a water pump operation status evaluation coefficient. The specific analysis formula for the water pump operation status evaluation coefficient is:

[0020]

[0021] where Eh represents the active power of the water pump, E represents the shaft power, Eh / E represents the water pump efficiency, Tp represents the historical operation time of the water pump, Tg represents the historical maintenance time of the water pump, u1 and u2 are the quadratic and linear coefficients of the historical maintenance time, respectively, and u1 > 0, u2 > 0; based on the intelligent analysis model of water pump performance, the water pump flow stability evaluation coefficient, the water pump pressure safety evaluation coefficient, and the water pump operation status evaluation coefficient are integrated, connected to the digital twin model of the water pump, and the physical behavior of the water pump is simulated.

[0022] Preferably, the water pump knowledge graph integrates the pre-processed water pump flow data, water pump pressure data, and water pump operation data based on the pre-processed water pump performance data to obtain a complete water pump performance dataset; according to the water pump performance dataset, the water pump flow data and water pump pressure data are classified based on the water pump operation data to obtain the analysis key points corresponding to the water pump performance dataset. The analysis key points corresponding to the water pump performance dataset are the water pump knowledge fields corresponding to the water pump performance dataset. The water pump knowledge fields include, but are not limited to, the water pump flow analysis field, the water pump pressure analysis field, and the water pump operation analysis field. Different water pump knowledge graphs are constructed according to different water pump knowledge fields.

[0023] Preferably, S5 combines with the water pump digital twin model to determine the water pump operation state and perform visualization processing on the water pump. The water pump operation state includes the green zone, the yellow zone, and the red zone. If the water pump operation state enters the green zone, it means that the water pump is in a normal and safe operation state, and the system operates normally without additional monitoring or intervention; if the water pump operation state enters the yellow zone, it means that the water pump is in a potential risk or slightly abnormal operation state, and the system is about to have problems or already has some slight abnormalities. At this time, a warning prompt is sent through the twin platform so that the operator can pay attention in time and take necessary preventive measures; if the water pump operation state enters the red zone, it means that the water pump is in a severely abnormal or dangerous operation state, and the system has already had problems or is about to have serious failures. At this time, the twin platform prompts to handle according to the pre-plan to ensure the safety and stability of the system; the pre-plan formulates corresponding handling pre-plans through the pre-plan model according to the reasons for entering the "red zone" during operation, such as cavitation, pressure pulsation, hump, and over-water level. The method for judging whether the water pump enters the red zone is as follows:

[0024] D1: Obtain real-time perception data from physical entities. The physical entities include a monitoring system and a data processing system. The real-time perception data includes the water pump performance data collected and pre-processed.

[0025] D2: In the virtual body, first, digitize the performance curve into a digital model; secondly, according to the obtained real-time perception data or manually input the corresponding data, start the decision support algorithm, calculate the unit number and the corresponding speed to be operated, and output the calculation result to the database.

[0026] D3: Synchronously display the virtual operating point trajectory and the real-time operating point on the virtual screen.

[0027] D4: The virtual operating point is corrected according to historical big data.

[0028] D5: When the operating point enters the "red zone", propose corresponding handling pre-plans.

[0029] Preferably, the visualization processing includes a user interface layer and a system service layer. Through the user interface, using front-end development technologies, functions such as measuring point management, instrument management, 3D view, and monitoring data query are realized, and an interactive interface is set to allow users to interact with the digital twin model of the water pump, such as adjusting parameters and viewing historical data. Through the system service layer, common service functions are provided to generate detailed data reports and analysis reports for management personnel to use.

[0030] Preferably, S6 optimizes the digital twin of the water pump in a modular manner based on the water pump performance dataset, the water pump performance intelligent analysis model, and the water pump operating state. The modularization includes a data module, a model module, and an application module. The data module, the model module, and the application module are connected and communicate through reserved interface slots. The reserved interface slots adopt a microservices architecture or a plug-in architecture to add a mechanism model to the digital twin when needed. The architecture supports dynamic loading and unloading of modules to adapt to different application scenarios and requirements. The data module is used to collect water pump performance data and continuously integrate the water pump performance dataset, establish a real-time mapping mechanism, perform real-time synchronization and update of data between the physical entity and the digital twin of the water pump, and provide an interface for data exchange with other modules. Through the reserved interface slots, new sensors or data sources are added to obtain new sensed data, and the newly added sensed data is automatically integrated into the water pump performance dataset in the data module to update the digital twin of the water pump in real time. The model module optimizes the water pump performance intelligent analysis model based on the water pump performance dataset, adjusts the parameters, optimizes the structure, and retrains the model according to actual needs and data changes. The water pump performance intelligent analysis model includes, but is not limited to, a water pump flow digital model, a water pump pressure digital model, and a water pump operation digital model. And based on the water pump flow digital model, the water pump pressure digital model, the water pump operation digital model, and the updated mechanism model, the rationality of the digital twin of the water pump is calculated. The specific calculation formula is:

[0031]

[0032] Where P represents the rationality of the digital twin of the water pump, Y represents the water pump flow stability evaluation coefficient, Z represents the water pump pressure safety evaluation coefficient, T represents the water pump operation state evaluation coefficient, λ1, λ2, and λ3 are proportionality coefficients corresponding to different evaluation dimensions, λ1>0, λ2>0, λ3>0, φ is the rationality adjustment parameter of the updated mechanism model, and the rationality P of the digital twin of the water pump is compared with the preset rationality P of the digital twin of the water pump yuMake a comparison to determine whether the digital twin of the water pump is reasonable, give an early warning to the unreasonable digital twin of the water pump and further optimize and update it; the data model also provides interfaces for model calling and data transfer with other modules, and updates or replaces the mechanism model in the existing model module through reserved interface slots to adapt to new physical entities and changing environmental conditions, and the updated mechanism model is seamlessly integrated into the digital twin of the water pump; the application module is based on the data module and the model module, monitors, gives early warnings, makes decisions and diagnoses faults for the use of the water pump according to the digital twin of the water pump, and drives the behavior parameters of the digital twin of the water pump through three methods: static values, simulated values and mechanism model data. Among them, the static value is used to initialize the behavior parameters of the digital twin of the water pump as the initial parameters, the simulated value is used when the digital twin of the water pump lacks specific data, and the mechanism model data is mostly used to drive simulation. The behavior parameters of the digital twin of the water pump include the control of the water outlet flow of the water pump, the control of the water head of the water pump, and the start and stop of the impeller rotation. Interact with the user according to different behavior parameters, apply the scenarios under different behavior parameters, and provide interfaces for integration with other systems or modules.

[0033] Technical effects and advantages of the present invention:

[0034] 1. By constructing a physical model, it provides a basis for the digital twin model of the water pump, collects the performance data of the water pump based on the physical model, and performs cleaning, denoising and normalization processing on the collected data to eliminate outliers and noise preprocessing in the data, improve the quality and reliability of the data, and also provide an important basis for the subsequent construction of the digital twin model of the water pump; by constructing the digital twin model of the water pump from multiple dimensions of the water pump flow data, water pump pressure data and water pump operation data, and continuously improving the model according to the actual operation of the water pump, it enables real-time monitoring of multiple dimensions of the water pump equipment, timely detection of abnormal situations, prediction of fault risks, and giving maintenance suggestions, which can significantly reduce the fault downtime, improve the stability and reliability of the equipment, analyze and predict the operation status of the water pump, so as to optimize the equipment operation and reduce the energy consumption cost;

[0035] 2. Based on the digital twin model of the water pump, through 3D visualization technology, the digital twin model of the water pump is presented to the management personnel in an intuitive and vivid way, which helps the management personnel better understand the operation status, performance parameters and potential problems of the water pump, adopts modular design, open data interfaces, etc., so as to facilitate subsequent function expansion and system integration, and realize the scalability of the digital twin model of the water pump; the digital twin model of the water pump can also conduct safety monitoring on the water resource allocation project, judge different operating conditions of the water pump, and make different response results, further improving the safety of the project. Description of the Drawings

[0036] Figure 1This is the overall structural schematic diagram of the present invention.

[0037] Figure 2 This is the schematic diagram for judging the red zone of the operating state of the water pump of the present invention. Specific embodiments

[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0039] As shown in the attached Figure 1 An expandable method for constructing a digital twin of a water pump shown, includes the following steps: S1 to S6.

[0040] S1: Construct a physical model: Construct a physical model according to the water pump required for the water resources allocation project.

[0041] S2: Collect water pump performance data: Based on the physical model constructed in S1, collect water pump performance data for the water resources allocation project. The water pump performance data includes water pump flow data, water pump pressure data, and water pump operation data.

[0042] In this embodiment, it should be specifically noted that S2 includes a water pump flow data collection unit, a water pump pressure data collection unit, and a water pump operation data collection unit. The water pump flow data collection unit is used to collect water pump flow data, including inlet flow, outlet flow, volume flow, and mass flow; the water pump pressure data collection unit is used to collect water pump pressure data, including inlet pressure, outlet pressure, and head; the water pump operation data collection unit is used to collect water pump operation data, including the historical operation time, historical maintenance time, active power of the water pump, and shaft power.

[0043] Furthermore, the outlet flow and inlet flow of the water pump are measured in real time by installing a flow volume flowmeter and a flow sensor at the water outlet and water inlet of the water pump respectively. The volume flow is the volume of liquid passing through the water pump outlet per unit time, and the mass flow is the mass of liquid passing through the water pump outlet per unit time; the outlet pressure of the water pump is the pressure value at the water pump outlet, representing the water flow pressure output by the water pump, and the inlet pressure of the water pump is the pressure value at the water pump inlet, representing the water flow pressure input by the water pump. A pressure gauge is used to measure by connecting to the inlet or outlet of the water pump. The head is the ability required for the water pump to pump water from a low place to a high place, and it is an important indicator to measure the liquid lifting ability of the water pump.

[0044] S3: Preprocess the pump performance data: Preprocess the pump performance data collected in S2, including data cleaning, data correction, filtering, and feature extraction.

[0045] In this embodiment, it should be specifically noted that the preprocessing automatically screens and eliminates abnormal data in the pump performance data, and uses data cleaning technology to correct the data; data cleaning refers to discovering and correcting identifiable errors in the data, including checking data consistency, handling invalid values and missing values. Filtering is to remove noise and interference in the pump performance data in the virtual body. Feature extraction is to extract information from the pump performance data collected in S2, including but not limited to the following features: Average value: Calculate the average historical maintenance time and average historical operation time over a period of time to reflect the operating state of the pump; Maximum and minimum values: Calculate the maximum and minimum outlet pressures and inlet pressures over a period of time to reflect the extreme working state of the pump; Percentage: Calculate the percentages of the pump, hydraulics, and volume over a period of time to reflect the efficiency of the pump.

[0046] S4: Build a pump digital twin model: Based on the pump performance data and physical model preprocessed in S3, establish a pump performance intelligent analysis model, and build a pump digital twin model according to the pump performance intelligent analysis model.

[0047] In this embodiment, it should be specifically noted that the pump digital twin model simulates based on the preprocessed pump performance data, creates a virtual modeling corresponding to the physical world, establishes a pump performance intelligent analysis model, and establishes a pump performance intelligent analysis model based on the pump knowledge graph. The pump performance intelligent analysis model includes but not limited to the pump flow digital model, the pump pressure digital model, and the pump operation digital model. The pump flow digital model is used to call the pump flow data in the pump performance data set and analyze the pump flow data to obtain the pump flow stability evaluation coefficient. The specific analysis formula for the pump flow stability evaluation coefficient is:

[0048]

[0049] Where Y represents the pump flow stability evaluation coefficient, Co represents the pump outlet flow, Ci represents the pump inlet flow, Co / Ci represents the pump volumetric efficiency, Cv represents the pump volume flow, Cm represents the pump mass flow, Cv 预 represents the preset volume of liquid passing through the pump outlet per unit time, Cm 预 represents the preset mass of liquid passing through the pump outlet per unit time; The pump pressure digital model is used to call the pump pressure data in the pump performance data set and analyze the pump pressure data to obtain the pump pressure safety evaluation coefficient. The specific analysis formula for the pump pressure safety evaluation coefficient is:

[0050]

[0051] Among them, Z represents the pump pressure safety evaluation coefficient, Lc represents the head of the pump, and Lc 理 represents the theoretical head of the pump, and Lc / Lc 理 represents the pump hydraulic efficiency, Po represents the outlet pressure of the pump, Pi represents the inlet pressure of the pump, and Po min represents the minimum value of the outlet pressure, and Po max represents the maximum value of the outlet pressure; the pump operation digital model is used to call the pump operation data in the pump performance data set and analyze the pump operation data to obtain the pump operation state evaluation coefficient. The specific analysis formula for the pump operation state evaluation coefficient is:

[0052]

[0053] Among them, Eh represents the active power of the pump, E represents the shaft power, Eh / E represents the pump efficiency, Tp represents the historical operation time of the pump, Tg represents the historical maintenance time of the pump, and u1 and u2 are the quadratic term coefficient and the linear term coefficient of the historical maintenance time, respectively, with u1 > 0 and u2 > 0;

[0054] Based on the pump performance intelligent analysis model, integrate the pump flow stability evaluation coefficient, the pump pressure safety evaluation coefficient, and the pump operation state evaluation coefficient, connect to the pump digital twin model, and simulate the pump physical behavior; the pump knowledge graph integrates the preprocessed pump flow data, pump pressure data, and pump operation data based on the preprocessed pump performance data to obtain a complete pump performance data set; according to the pump performance data set, classify the pump flow data and pump pressure data based on the pump operation data to obtain the analysis key points corresponding to the pump performance data set. The analysis key points corresponding to the pump performance data set are the pump knowledge fields corresponding to the pump performance data set. The pump knowledge fields include, but are not limited to, the pump flow analysis field, the pump pressure analysis field, and the pump operation analysis field. Different pump knowledge graphs are constructed according to different pump knowledge fields.

[0055] S5: Apply the pump digital twin model: Based on the pump digital twin model constructed in S4, apply the pump digital twin model and perform visualization processing.

[0056] In this embodiment, it should be specifically noted that in S5, in combination with the digital twin model of the water pump, the operating state of the water pump is determined, and visualization processing is performed on the water pump. The operating state of the water pump includes a green zone, a yellow zone, and a red zone. If the operating state of the water pump enters the green zone, it indicates that the water pump is in a normal and safe operating state, and the system operates normally without the need for additional monitoring or intervention. If the operating state of the water pump enters the yellow zone, it indicates that the water pump is in a potentially risky or slightly abnormal operating state, and the system is about to have problems or already has some minor abnormalities. At this time, a warning prompt is issued through the twin platform so that the operator can pay attention in time and take necessary preventive measures. If the operating state of the water pump enters the red zone, it indicates that the water pump is in a severely abnormal or dangerous operating state, and the system has already had problems or is about to have serious failures. At this time, the twin platform prompts to handle according to the pre-plan to ensure the safety and stability of the system. The pre-plan is formulated through a pre-plan model according to the reasons for entering the "red zone" during operation, such as cavitation, pressure pulsation, hump, and over-water level. The method for judging whether the water pump enters the red zone is as follows:

[0057] A1: Obtain real-time perception data from the physical entity. The physical entity includes a monitoring system and a data processing system. The real-time perception data includes the collected and preprocessed water pump performance data;

[0058] A2: In the virtual entity, first, digitize the performance curve into a digital model. Second, according to the obtained real-time perception data or manually input the corresponding data, start the decision support algorithm, calculate the unit number and corresponding speed to be operated, and output the calculation results to the database;

[0059] A3: Synchronously display the virtual operating point trajectory and the real-time operating point on the virtual screen;

[0060] A4: Correct the virtual operating point according to historical big data;

[0061] A5: When the operating point enters the "red zone", propose corresponding handling pre-plans.

[0062] Furthermore, the visualization processing includes a user interface layer and a system service layer. Through the user interface, using front-end development technologies, functions such as measuring point management, instrument management, 3D view, and monitoring data query are realized, and an interactive interface is set to allow users to interact with the digital twin model of the water pump, such as adjusting parameters and viewing historical data. Through the system service layer, common service functions are provided to generate detailed data reports and analysis reports for management personnel to use.

[0063] S6: Optimize the digital twin of the water pump: Based on the water pump performance data set, the water pump performance intelligent analysis model, and the operating state of the water pump, modularize and optimize the digital twin of the water pump.

[0064] In this embodiment, it should be specifically noted that in S6, based on the water pump performance dataset, the water pump performance intelligent analysis model, and the water pump operation status, modularization is used to optimize the water pump digital twin. The modularization includes a data module, a model module, and an application module. The data module, the model module, and the application module are connected and communicate through reserved interface slots. The reserved interface slots adopt a microservices architecture or a plug-in architecture to facilitate the addition of a mechanism model to the digital twin when needed. The architecture supports dynamic loading and unloading of modules to adapt to different application scenarios and requirements. The data module is used to collect water pump performance data, continuously integrate the water pump performance dataset, and provide an interface for data exchange with other modules. Through the reserved interface slots, new sensors or data sources are added to obtain new sensed data. The newly added sensed data is automatically integrated into the water pump performance dataset in the data module to update the water pump digital twin in real time. The model module optimizes the water pump performance intelligent analysis model based on the water pump performance dataset, adjusts the model parameters, optimizes the structure, and retrains the model according to actual requirements and data changes. The water pump performance intelligent analysis model includes, but is not limited to, a water pump flow digital model, a water pump pressure digital model, and a water pump operation digital model. The rationality of the water pump digital twin is calculated based on the water pump flow digital model, the water pump pressure digital model, the water pump operation digital model, and the updated mechanism model. The specific calculation formula is:

[0065]

[0066] where P represents the rationality of the water pump digital twin, Y represents the water pump flow stability evaluation coefficient, Z represents the water pump pressure safety evaluation coefficient, T represents the water pump operation status evaluation coefficient, λ1, λ2, and λ3 are proportionality coefficients corresponding to different evaluation dimensions, λ1>0, λ2>0, λ3>0, φ is the rationality adjustment parameter of the updated mechanism model, and the rationality P of the water pump digital twin is compared with the preset rationality P of the water pump digital twin yuMake a comparison to determine whether the digital twin of the water pump is reasonable, give early warnings to the unreasonable digital twin of the water pump, and further optimize and update it; the data model also provides interfaces for model calls and data transmission with other modules, and updates or replaces the mechanism model in the existing model module through reserved interface slots to adapt to new physical entities and changing environmental conditions. The updated mechanism model is seamlessly integrated into the digital twin of the water pump; the application module is based on the data module and the model module, monitors, gives early warnings, makes decisions, and diagnoses faults for the use of the water pump according to the digital twin of the water pump. Through three methods of static values, simulated values, and mechanism model data, the behavior parameters of the digital twin of the water pump are driven. Among them, the static value is used to initialize the behavior parameters of the digital twin of the water pump as the initial parameters, the simulated value is used when the digital twin of the water pump lacks specific data, and the mechanism model data is mostly used to drive simulation. The behavior parameters of the digital twin of the water pump include the control of the water outlet flow of the water pump, the control of the water head of the water pump, and the start and stop of the impeller rotation. Interact with users according to different behavior parameters, apply scenarios under different behavior parameters, and provide interfaces for integration with other systems or modules.

[0067] The present invention provides a basis for the digital twin model of the water pump by constructing a physical model, collects the performance data of the water pump based on the physical model, and performs cleaning, denoising, and normalization processing on the collected data to eliminate outliers and noise in the data for preprocessing, improve the quality and reliability of the data, and also provide an important basis for the subsequent construction of the digital twin model of the water pump; construct the digital twin model of the water pump from multiple dimensions of the water pump flow data, water pump pressure data, and water pump operation data, and continuously improve the model according to the actual operation of the water pump, so as to monitor multiple dimensions of the water pump equipment in real time, timely detect abnormal situations, predict the risk of failure, and give maintenance suggestions, which can significantly reduce the failure downtime, improve the stability and reliability of the equipment, analyze and predict the operation state of the water pump, so as to optimize the equipment operation and reduce the energy consumption cost; based on the digital twin model of the water pump, through three-dimensional visualization technology, display the digital twin model of the water pump to the management personnel in an intuitive and vivid way, which helps the management personnel better understand the operation state, performance parameters, and potential problems of the water pump, adopt modular design, open data interfaces, etc., so as to facilitate subsequent function expansion and system integration, and realize the scalability of the digital twin model of the water pump; the digital twin model of the water pump can also perform safety monitoring on the water resource allocation project, judge different operating conditions of the water pump, and make different response results, further improving the safety of the project.

[0068] Secondly: In the attached drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments of the present disclosure are involved, and other structures can refer to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other;

[0069] Finally, the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An extensible method for constructing a digital twin of a water pump, characterized in that Including: S1: Construct a physical model: Construct a physical model according to the water pumps required for the water resources allocation project; S2: Collect pump performance data: Based on the physical model constructed in S1, collect pump performance data for the water resources allocation project. The pump performance data includes pump flow rate data, pump pressure data, and pump operation data; S3: Preprocess pump performance data: Preprocess the pump performance data collected in S2, including data cleaning, data correction, filtering, and feature extraction; S4: Construct a pump digital twin model: Establish a pump performance intelligent analysis model based on the pump performance data and physical model preprocessed in S3, and construct a pump digital twin model according to the pump performance intelligent analysis model; The pump digital twin model is simulated based on the preprocessed pump performance data, creating a virtual modeling corresponding to the physical world, establishing a pump performance intelligent analysis model, and establishing a pump performance intelligent analysis model based on the pump knowledge graph. The pump performance intelligent analysis model includes, but is not limited to, a pump flow rate digital model, a pump pressure digital model, and a pump operation digital model. The pump flow rate digital model is used to retrieve the pump flow rate data in the pump performance data set and analyze the pump flow rate data to obtain a pump flow rate stability evaluation coefficient. The specific analysis formula for the pump flow rate stability evaluation coefficient is: ， Where Y represents the pump flow stability evaluation coefficient, Co represents the pump outlet flow rate, Ci represents the pump inlet flow rate, Co / Ci represents the pump volumetric efficiency, Cv represents the pump volume flow rate, Cm represents the pump mass flow rate, Cv 预 represents the preset volume of liquid passing through the pump outlet per unit time, Cm 预 represents the preset mass of liquid passing through the pump outlet per unit time; The pump pressure digital model is used to retrieve the pump pressure data in the pump performance data set and analyze the pump pressure data to obtain a pump pressure safety evaluation coefficient. The specific analysis formula for the pump pressure safety evaluation coefficient is: ， Where Z represents the safety assessment coefficient of the pump pressure, Lc represents the head of the pump, and Lc 理 represents the theoretical head of the pump, and Lc / Lc 理 represents the hydraulic efficiency of the pump, Po represents the outlet pressure of the pump, Pi represents the inlet pressure of the pump, and Po min represents the minimum value of the outlet pressure, and Po max represents the maximum value of the outlet pressure; The pump operation digital model is used to retrieve the pump operation data in the pump performance data set and analyze the pump operation data to obtain a pump operation status evaluation coefficient. The specific analysis formula for the pump operation status evaluation coefficient is: ， Where, T represents the pump operation status evaluation coefficient, Eh represents the active power of the pump, E represents the shaft power, Eh / E represents the pump efficiency, Tp represents the historical operation time of the pump, Tg represents the historical maintenance time of the pump, u1 and u2 are the quadratic term coefficient and the linear term coefficient of the historical maintenance time, and u1>0, u2>0; Based on the pump performance intelligent analysis model, integrate the pump flow rate stability evaluation coefficient, the pump pressure safety evaluation coefficient, and the pump operation status evaluation coefficient, connect to the pump digital twin model, and simulate the pump physical behavior; S5: Apply the pump digital twin model: Based on the pump digital twin model constructed in S4, apply the pump digital twin model and perform visualization processing; S6: Optimize the pump digital twin: Optimize the pump digital twin modularly based on the pump performance data set, the pump performance intelligent analysis model, and the pump operation status.

2. The method for constructing an extensible digital twin of a water pump according to claim 1, characterized in that: The S2 includes a pump flow rate data collection unit, a pump pressure data collection unit, and a pump operation data collection unit. The pump flow rate data collection unit is used to collect pump flow rate data, including inlet flow rate, outlet flow rate, volume flow rate, and mass flow rate; The water pump pressure data acquisition unit is used to acquire water pump pressure data, including inlet pressure, outlet pressure, and head; the water pump operation data acquisition unit is used to acquire water pump operation data, including the historical operation time, historical maintenance time, active power of the water pump, and shaft power.

3. A method for constructing an extensible digital twin of a water pump according to claim 1, characterized in that: The preprocessing automatically screens and eliminates abnormal data in the water pump performance data, and uses data cleaning technology to correct the data; data cleaning refers to discovering and correcting identifiable errors in the data, including checking data consistency, handling invalid values and missing values. Filtering is to remove noise and interference in the water pump performance data in the virtual body. Feature extraction is to extract information from the water pump performance data collected in S2, including but not limited to the following aspects of features: Average value: Calculate the average historical maintenance time and average historical operation time over a period of time to reflect the operation status of the water pump; Maximum and minimum values: Calculate the maximum and minimum outlet pressures and inlet pressures over a period of time to reflect the extreme working state of the water pump; Percentage: Calculate the percentages of the water pump, hydraulic, and volume over a period of time to reflect the efficiency of the water pump.

4. A method for constructing an extensible digital twin of a water pump according to claim 1, characterized in that: The water pump knowledge graph integrates the preprocessed water pump flow data, water pump pressure data, and water pump operation data based on the preprocessed water pump performance data to obtain a complete water pump performance data set; according to the water pump performance data set, classify the water pump flow data and water pump pressure data based on the water pump operation data to obtain the analysis key points corresponding to the water pump performance data set. The analysis key points corresponding to the water pump performance data set are the water pump knowledge fields corresponding to the water pump performance data set. The water pump knowledge fields include but are not limited to the water pump flow analysis field, the water pump pressure analysis field, and the water pump operation analysis field. Different water pump knowledge graphs are constructed according to different water pump knowledge fields.

5. The method for constructing an extensible digital twin of a water pump according to claim 1, characterized in that: S5 combines the water pump digital twin model to determine the water pump operation status and perform visualization processing on the water pump. The water pump operation status includes the green zone, the yellow zone, and the red zone. If the water pump operation status enters the green zone, it means that the water pump is in a normal and safe operation state, and the system operates normally without additional monitoring or intervention; if the water pump operation status enters the yellow zone, it means that the water pump is in a potentially risky or slightly abnormal operation state, and the system is about to have problems or already has some minor abnormalities. At this time, a warning prompt is sent through the twin platform so that the operator can pay attention in time and take necessary preventive measures; if the water pump operation status enters the red zone, it means that the water pump is in a severely abnormal or dangerous operation state, and the system has already had problems or is about to have a serious failure. At this time, the twin platform prompts to handle according to the pre-plan to ensure the safety and stability of the system.

6. The method for constructing an extensible digital twin of a water pump according to claim 5, characterized in that: The visualization processing includes a user interface layer and a system service layer. Through the user interface and using front-end development technologies, functions such as measurement point management, instrument management, 3D view, and monitoring data query are realized, and an interactive interface is set to allow users to interact with the water pump digital twin model, such as adjusting parameters and viewing historical data. The system service layer provides common service functions to generate detailed data reports and analysis reports for management personnel to use.

7. A method for constructing an extensible digital twin of a water pump according to claim 1, characterized in that: The S6 optimizes the water pump digital twin by modularization based on the water pump performance dataset, the water pump performance intelligent analysis model, and the water pump operation status. The modularization includes a data module, a model module, and an application module. The data module, the model module, and the application module are connected and communicate through reserved interface slots. The reserved interface slots adopt a microservices architecture or a plug-in architecture to add mechanism models to the digital twin when needed. The architecture supports dynamic loading and unloading of modules to adapt to different application scenarios and requirements. The data module is used to collect the pump performance data, continuously integrate the pump performance data set, establish a real-time mapping mechanism, perform real-time synchronization and update of the data between the physical entity and the pump digital twin, and provide an interface for data exchange with other modules. By reserving interface slots, new sensors or data sources can be added to obtain new sensed data. The newly added sensed data is automatically integrated into the pump performance data set in the data module to perform real-time update of the pump digital twin. The model module optimizes the pump performance intelligent analysis model based on the pump performance data set. According to the actual requirements and data changes, the parameters of the model are adjusted, the structure is optimized, and retraining is performed. The pump performance intelligent analysis model includes, but is not limited to, the pump flow digital model, the pump pressure digital model, and the pump operation digital model. And based on the pump flow digital model, the pump pressure digital model, the pump operation digital model, and the updated mechanism model, the rationality of the pump digital twin is calculated. The specific calculation formula is: , Among them, P represents the rationality of the pump digital twin, Y represents the evaluation coefficient of the pump flow stability, Z represents the evaluation coefficient of the pump pressure safety, T represents the evaluation coefficient of the pump operation state, λ1, λ2, and λ3 are the proportionality coefficients corresponding to different evaluation dimensions, λ1>0, λ2>0, λ3>0, φ is the adjustment parameter of the updated mechanism model rationality. Compare the rationality P of the pump digital twin with the preset rationality P of the pump digital twin yu to judge whether the pump digital twin is reasonable, give early warnings to the unreasonable pump digital twins and further optimize and update them; the data model also provides interfaces for model calling and data transfer with other modules, and updates or replaces the mechanism model in the existing model module through reserved interface slots to adapt to new physical entities and changing environmental conditions. The updated mechanism model is seamlessly integrated into the pump digital twin; Based on the data module and the model module, the application module monitors, warns, makes decisions, and diagnoses faults for the use of the water pump according to the water pump digital twin. The behavior parameters of the water pump digital twin are driven through three methods: static values, simulation values, and mechanism model data. Among them, static values are used to initialize the behavior parameters of the water pump digital twin as initial parameters. Simulation values are used when the water pump digital twin lacks specific data. Mechanism model data is mostly used to drive simulation. The behavior parameters of the water pump digital twin include water pump outlet flow control, water pump head control, and impeller rotation start and stop. Interact with users according to different behavior parameters, apply scenarios under different behavior parameters, and provide interfaces for integration with other systems or modules.

Images