Method and system for monitoring operation state of water pump unit based on digital twin technology

Abstract

The invention relates to a method and system for monitoring the running state of a water pump unit based on the digital twin technology, and belongs to the technical field of pipeline pump monitoring. The method comprises the steps that a digital twin model fusing equipment parameters and environment sensing data is constructed, and multi-dimensional running data such as dynamic pressure, flow, electric energy quality and temperature are collected in real time in combination with the Internet of Things technology; a multi-dimensional correction algorithm is innovatively provided, fluid flow velocity calculation is corrected through pipe opening structure parameters, lift calculation is dynamically corrected through a temperature compensation algorithm and a mechanical wear factor, and dynamic compensation is conducted on the water pump efficiency by introducing a harmonic loss calibration mechanism. Based on a two-channel machine learning model, a fault incidence matrix is constructed by improving a weighted K-means algorithm to realize equipment health state clustering, an efficiency attenuation trend is predicted in combination with an LSTM network, static characteristics and dynamic time sequence data are fused by using an attention mechanism, and finally an accurate fault diagnosis report and a maintenance strategy are generated.

Description

Technical Field

[0001] This application relates to the technical field of pipeline pump monitoring, and more specifically, to a method and system for monitoring the operating status of a water pump unit based on digital twin technology. Background Art

[0002] As the core link of urban water supply, the operating efficiency of water pumps in water plants is directly related to the water supply cost, energy consumption, and water supply quality. In recent years, with the rapid development of automation technology, sensor technology, and information technology, great progress has been made in the technology for monitoring the operating efficiency of water pumps in water plants. However, there are still significant technical bottlenecks in actual working condition monitoring: ① The traditional monitoring system adopts a distributed architecture, and independent subsystems such as flow meters, watt-hour meters, and vibration sensors form data islands, resulting in a timing deviation between the pipe loss compensation factor and the motor input power in the head calculation; ② Existing digital twin applications mostly focus on single physical field simulation, lacking the ability of electromechanical coupling modeling and unable to analyze the non-linear impact of harmonic distortion on motor efficiency; ③ The operation and maintenance mode based on threshold alarm is difficult to capture progressive faults such as impeller cavitation and bearing wear.

[0003] The prior art, such as the Chinese patent application with the publication number CN118309644A, discloses a method for monitoring the operating flow rate of a pipeline pump based on digital twin. The method includes: a data acquisition layer, a communication network layer, a data processing layer, a digital twin model layer, a flow rate monitoring and prediction layer, a decision support and control layer, and a user interface layer. By constructing a digital twin model of the pipeline pump and combining with the real-time monitoring of the flow rate, the invention realizes the real-time monitoring, intelligent prediction, precise control, and efficient management of the pipeline pump, thus effectively improving the operating efficiency, stability, and economic benefits of the device, and providing strong support for improving the performance and benefits of the pipeline pump system.

[0004] The problems of the above prior art are that through the monitoring of the flow rate, the operating output of the water pump can be obtained, reflecting whether the operating state of the water pump is normal, but the energy consumption data of the water pump is not collected, the input power of the water pump cannot be analyzed and calculated, and the evaluation of the operating efficiency of the water pump and the analysis ability of potential faults are lacking. Therefore, it is difficult to provide data support for effective water pump operation regulation and fault troubleshooting to maintain the water pump in the best efficiency operating state. There is a lack of a dynamic parameter correction mechanism, and the dynamic influence of environmental parameters (such as water temperature, pipeline thermal expansion) and equipment degradation (such as impeller wear) on the monitoring results is not considered. Summary of the Invention

[0005] To solve the above technical problems, the present invention proposes a method and system for monitoring the operating status of a water pump unit based on digital twin technology.

[0006] The technical solution of the present invention is as follows: The present invention provides a method for monitoring the operating status of a water pump unit based on digital twin technology, including the following steps: Construct a digital twin model of the water pump unit, which integrates the equipment parameters and environmental sensing parameters of the water pump unit; Obtain the operating status data of the water pump unit in real time through Internet of Things sensors. The water pump operating status data includes dynamic pressure values, instantaneous flow rates, power quality parameters, and fluid temperature parameters; Based on the digital twin model and the real-time operating status data of the water pump unit, dynamically calculate the actual operating efficiency of the water pump unit and perform multi-dimensional correction; Combining the historical operating status data of the water pump unit and the corrected actual operating efficiency, conduct fault diagnosis and efficiency evaluation of the water pump unit through a machine learning model to obtain a fault diagnosis and evaluation report of the water pump unit.

[0007] As a preferred embodiment, the equipment parameters include: a water pump geometric feature parameter group and a metering device feature parameter group, where: The water pump geometric feature parameter group includes the inner diameters of the inlet and outlet water pipes of the pump body and the baseline efficiency of the motor; The metering device feature parameter group includes the pulse constant of the electric energy meter, the transformation ratio of the voltage transformer, and the installation elevation of the pressure sensor.

[0008] As a preferred embodiment, the multi-dimensional correction includes applying the nozzle structure parameters to correct the calculation of the fluid velocity, specifically: ; ; In the formula, is the inlet velocity; is the outlet velocity; Q is the flow rate; is the inner diameter of the inlet pipe; is the inner diameter of the outlet pipe.

[0009] As a preferred embodiment, the multi-dimensional correction includes using a temperature compensation algorithm and a mechanical wear factor to dynamically correct the calculation of the head, specifically: ; Where: ; In the formula, H is the head; is the outlet pressure of the water pump; is the inlet pressure of the water pump; is the liquid density at the standard working condition water temperature; g is the acceleration due to gravity; T is the real-time water temperature; is the standard working condition water temperature; is the pipeline thermal expansion coefficient; is the impeller wear coefficient; is the elevation of the pressure gauge at the pump outlet; is the elevation of the pressure gauge at the pump inlet; is the cumulative operation time of the impeller; is the impeller wear acceleration threshold; is the impeller attenuation rate.

[0010] As a preferred embodiment, the multi-dimensional correction includes introducing power quality parameters to dynamically calibrate the actual operation efficiency of the pump, including the following process: Calculate the real-time effective power of the pump. The specific formula is: ; In the formula, is the real-time effective power of the pump; is the liquid density at the real-time water temperature; Calculate the real-time input power of the pump. The specific formula is: ; In the formula, N is the real-time input power of the pump; is the voltage transformer ratio; is the electric energy meter pulse constant; is the time for 10 revolutions of the electric energy meter; Calculate the comprehensive efficiency of the pump. The specific formula is: ; In the formula, is the comprehensive efficiency of the pump; Calculate the actual operation efficiency of the pump. The specific formula is: ; In the formula, is the actual operation efficiency of the pump; is the motor baseline efficiency; is the real-time power factor; is the rated power factor; is the harmonic loss coefficient; THD is the current waveform distortion rate.

[0011] As a preferred embodiment, the machine learning model adopts a dual-channel heterogeneous architecture, including: Feature clustering channel: Generate the clustering center of the device health status based on the improved weighted K-means algorithm. The objective function of the improved weighted K-means algorithm is: ; In the formula: J is the objective function of the improved weighted K-means algorithm; C is the total number of clusters; l is the index of the l-th cluster; i is the pump sample index; is the time decay function; is the actual operating efficiency of the i-th water pump; is the clustering center of the actual operating efficiency of the l-th clustering cluster; is the production duration balance coefficient; is the production duration of the i-th water pump; is the clustering center of the production duration of the l-th clustering cluster; m is the total number of failure types; is the failure type index, representing the j-th failure type and the q-th failure type respectively; is the element of the failure correlation matrix; is the indication value of whether the i-th water pump has the q-th failure type; is the average occurrence frequency of the q-th failure in the l-th clustering cluster; Time series prediction channel: An efficiency decay trend prediction model is constructed using an LSTM network; the objective function of the LSTM network is to design a multi-task time series weighted loss function based on the LSTM cross-entropy loss function, specifically: ; In the formula, is the multi-task time series weighted loss function; is the classification cross-entropy loss; is the signal-to-noise ratio weight adjustment coefficient; is the phase compensation coefficient; is the dynamic decay weight; is the predicted efficiency value within the time segment t; is the actual efficiency value within the time segment t; is the wear balance factor; is the KL divergence calculation; is the real-time probability distribution of the d-th feature cluster; is the reference probability distribution of the d-th feature cluster; D is the number of feature clusters; d is the index of the feature cluster; is the time window length; Cross-channel fusion layer: The coupling analysis of the static features of the feature clustering channel and the dynamic time series features of the time series prediction channel is realized through the attention mechanism.

[0012] As a preferred implementation, the element of the failure correlation matrix has the following specific calculation formula: ; In the formula: is the historical number of times that failure type j and failure type q occur simultaneously; is the total number of times that failure type j occurs alone; is the total number of times that failure type q occurs alone; is a constant to prevent the denominator from being 0; .

[0013] On the other hand, the present invention also provides a system for monitoring the operating status of a water pump unit based on digital twin technology, including: A digital twin model creation module that constructs a digital twin model of the water pump unit, and the digital twin model integrates the equipment parameters and environmental sensing parameters of the water pump unit; A data acquisition module that obtains the operating status data of the water pump unit in real time through Internet of Things sensors, and the water pump operating status data includes dynamic pressure values, instantaneous flow rates, power quality parameters, and fluid temperature parameters; An efficiency calculation module that dynamically calculates the actual operating efficiency of the water pump unit based on the digital twin model and the real-time operating status data of the water pump unit, and performs multi-dimensional correction; An evaluation and diagnosis module that combines the historical operating status data of the water pump unit and the corrected actual operating efficiency, and performs fault diagnosis and efficiency evaluation of the water pump unit through a machine learning model to obtain a fault diagnosis and evaluation report of the water pump unit.

[0014] The present invention has the following beneficial effects: 1. By dynamically correcting the head calculation through the temperature compensation algorithm and the impeller wear index decay model, and calibrating in combination with power quality parameters, the calculation accuracy of the effective power and input power is significantly improved, and the actual efficiency error is reduced.

[0015] 2. An improved weighted K-means algorithm is used to construct a fault correlation matrix to quantify the probability of fault concurrency, and combined with the LSTM time series prediction model, the equipment degradation trend analysis and early fault warning are realized.

[0016] 3. Based on the digital twin visualization platform, the efficiency evolution trend is tracked, and the attention mechanism integrating static features and dynamic time series is used to generate targeted maintenance strategies, reduce energy consumption and extend the equipment life.

[0017] 4. Integrate the data of flow meters, pressure gauges, watt-hour meters and environmental sensors to construct a multi-dimensional electromechanical coupling model, break through the limitations of traditional subsystem data islands, and realize the global perception of the operating status of the water pump.

[0018] 5. The two-channel heterogeneous machine learning model reduces the complexity of LSTM through clustering preprocessing, and combines KL divergence and multi-task loss function optimization to improve the prediction robustness and diagnosis accuracy. Description of the Drawings

[0019] Figure 1 It is a schematic diagram of the method flow for Embodiment 1; Figure 2 It is a schematic diagram of the structure of the water pump and monitoring equipment. Detailed Embodiments

[0020] 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.

[0021] It should be understood that the step numbers used in the text are only for convenient description and do not limit the execution order of the steps.

[0022] It should be understood that the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0023] The terms "comprising" and "including" indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0024] The term " / or" refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] Embodiment 1: To make the purpose, technical solutions and advantages of the present invention clearer, the following will combine specific embodiments of the present application and refer to the attached Figure 1 , and clearly and completely describe the technical solutions of the present invention.

[0026] To solve the problems of the existing technology, the present invention provides a method for monitoring the operating state of a water pump unit based on digital twin technology, including the following steps: Construct a digital twin model of the water pump unit, and the digital twin model integrates the equipment parameters and environmental sensing parameters of the water pump unit; The equipment parameters include: a water pump geometric feature parameter group and a metering device feature parameter group, where: The water pump geometric feature parameter group includes the inner diameters of the inlet and outlet water pipes of the pump body and the baseline efficiency of the motor; The metering device feature parameter group includes the pulse constant of the electric energy meter, the transformation ratio of the voltage transformer, and the installation elevation of the pressure sensor.

[0027] Real-time obtain the operating state data of the water pump unit through Internet of Things sensors, and the water pump operating state data includes dynamic pressure values, instantaneous flow rates, power quality parameters and fluid temperature parameters; In this embodiment, the power quality parameters include a time-domain and frequency-domain composite parameter set, specifically including: Time-domain parameters: active power fluctuation coefficient, current waveform distortion rate; the active power fluctuation coefficient is used to evaluate power stability.

[0028] Frequency-domain parameters: characteristic harmonic component amplitude spectrum, fundamental power factor.

[0029] Based on the digital twin model and the real-time operation status data of the water pump unit, dynamically calculate the actual operation efficiency of the water pump unit and perform multi-dimensional correction; The multi-dimensional correction includes applying the pipe orifice structure parameters to correct the fluid flow velocity calculation, specifically: ; ; In the formula, is the inlet flow velocity; is the outlet flow velocity; Q is the flow rate; is the inner diameter of the inlet pipe; is the inner diameter of the outlet pipe.

[0030] The multi-dimensional correction includes using a temperature compensation algorithm and a mechanical wear factor to dynamically correct the head calculation, specifically: ; Among them: ; In the formula, H is the head; is the pump outlet pressure; is the pump inlet pressure; is the liquid density at the standard working condition water temperature; g is the acceleration due to gravity; T is the real-time water temperature, obtained through the temperature sensor embedded in the digital twin model; is the standard working condition water temperature, default taken as 20 °C; is the pipe thermal expansion coefficient, taken as 0.0002 / °C; is the impeller wear coefficient; is the elevation of the pump outlet pressure gauge; is the elevation of the pump inlet pressure gauge; is the cumulative operation time of the impeller; is the impeller wear acceleration threshold, taken as 8000 hours; is the impeller attenuation rate, taken as 0.003 according to the wear model parameter range of similar mechanical components; by introducing the wear acceleration threshold and the impeller attenuation rate , the defect that the traditional linear model cannot capture the sudden change of equipment performance is solved. For example: when , the impeller wears slowly ; when When the time is up, the impeller wear rate increases exponentially, rapidly approaches 0, which highly coincides with the three-stage law of actual impeller wear (running-in period - stable period - failure period).

[0031] The multi-dimensional correction includes introducing power quality parameters to dynamically calibrate the actual operating efficiency of the water pump, including the following process: Calculate the real-time effective power of the water pump. The specific formula is: ; In the formula, is the real-time effective power of the water pump; is the liquid density at the real-time water temperature; Calculate the real-time input power of the water pump. The specific formula is: ; In the formula, N is the real-time input power of the water pump; is the voltage transformer ratio; is the meter pulse constant; is the time for 10 revolutions of the meter; Calculate the comprehensive efficiency of the water pump. The specific formula is: ; In the formula, is the comprehensive efficiency of the water pump; Calculate the actual operating efficiency of the water pump. The specific formula is: ; In the formula, is the actual operating efficiency of the water pump; is the motor baseline efficiency; is the real-time power factor, measured by the embedded meter; is the rated power factor, obtained from the motor nameplate; is the harmonic loss coefficient, calibrated by the amplitude spectrum of the characteristic harmonic components; THD is the current waveform distortion rate, obtained by the harmonic analysis of the meter.

[0032] Combining the historical operating state data of the water pump unit and the corrected actual operating efficiency, conduct fault diagnosis and efficiency evaluation of the water pump unit through a machine learning model, and obtain a fault diagnosis and evaluation report of the water pump unit.

[0033] The machine learning model adopts a dual-channel heterogeneous architecture, including: Feature clustering channel: Generate the clustering center of the device health status based on the improved weighted K-means algorithm; Time series prediction channel: Construct an efficiency decay trend prediction model using the LSTM network; Cross-channel fusion layer: Through the attention mechanism, the coupling analysis of the static features of the feature clustering channel and the dynamic time-series features of the time-series prediction channel is realized.

[0034] In this embodiment, the feature clustering channel: Based on the improved weighted K-means algorithm, the clustering centers of the device health status are generated, which specifically includes the following processes: Collect the operation data of the historical water pump throughout its life cycle. The data includes: water pump model, actual operation efficiency of the water pump throughout its life cycle, water pump failure type and corresponding energy consumption, flow rate, and head. Classify the water pumps according to different failure types. The failure types are divided into m categories. Then the failure situation of the water pump is expressed by the following formula: F ; In the formula, F is the failure situation of the water pump; if the water pump has the jth failure, then , otherwise 0; Calculate the clustering center of each classification cluster; the specific calculation formula is: ; ; ; In the formula: is the clustering center of the production duration of the lth clustering cluster; is the clustering center of the actual operation efficiency of the lth cluster; is the jth component of the failure type clustering center; is the production duration of the ith water pump; is the actual operation efficiency of the ith water pump; is the ith water pump regarding the jth failure type; is the number of water pumps in the lth category; is the set belonging to the lth category of water pumps; ; The objective function of the improved weighted K-means algorithm in the feature clustering channel is: ; In the formula: J is the objective function of the improved weighted K-means algorithm; C is the total number of clustering clusters; l is the lth clustering cluster; i is the water pump sample index, indicating the ith water pump; is the time decay function; is the actual operation efficiency of the ith water pump; is the clustering center of the actual operation efficiency of the lth cluster; is the production duration balance coefficient; is the production duration of the ith water pump; is the clustering center of the production start-up duration of the l-th clustering cluster; m is the total number of failure types; are failure type indices, representing the j-th failure type and the q-th failure type respectively; is an element of the failure correlation matrix; is an indication value of whether the i-th water pump has the q-th failure type, 1 means yes, 0 means no; is the average occurrence frequency of the q-th failure in the l-th clustering cluster; Among them, the element of the failure correlation matrix, the specific calculation formula is: ; In the formula: is the historical number of times that failure type j and failure type q occur simultaneously; is the total number of times that failure type j occurs alone; is the total number of times that failure type q occurs alone; is a very small constant to prevent the denominator from being 0; ; Among them, the time decay function, which decreases with the increase of the production start-up time, the specific calculation formula is: ; In the formula: is the decay rate parameter.

[0035] Through iterative calculation, make the objective function J reach the minimum value, and obtain a reasonable clustering of the production start-up duration of the water pump, the actual operating efficiency of the water pump, and the water pump failure situation.

[0036] Time series prediction channel: Use an LSTM network to construct an efficiency decay trend prediction model; specifically include the following steps: Arrange the data of each classification set of the clustered water pumps in chronological order, and divide the data points of each set into a training set and a test set; Extract features for each classification arranged in chronological order to obtain a feature cluster, the specific formula is: ; In the formula, is the feature row vector of the q-th row of the p-th set; is the average value of the actual operating efficiency of the water pumps in the p-th set; is the standard deviation of the actual operating efficiency of the p-th set; is the average value of the production start-up time of the water pumps in the p-th set; is the standard deviation of the production start-up time of the water pumps in the p-th set; is the failure frequency of m different failure types in the p-th set; An efficiency decay trend prediction model is constructed by using a long short-term memory neural network (LSTM) of a time series prediction channel trained with a training set. The objective function of the LSTM network is to design a multi-task time series weighted loss function based on the LSTM cross-entropy loss function. Specifically, it is as follows: ; In the formula, is the multi-task time series weighted loss function; is the categorical cross-entropy loss; is the signal-to-noise ratio weight adjustment coefficient, ; is the phase compensation coefficient, ; is the dynamic decay weight; is the predicted efficiency value within the time segment t; is the actual efficiency value within the time segment t; is the wear leveling factor, = ; is the KL divergence calculation; is the real-time probability distribution of the d-th feature cluster; is the reference probability distribution of the d-th feature cluster; D is the number of feature clusters; d is the index of the feature cluster; is the time window length; Among them, the dynamic decay weight assigns higher weights to recent time steps. The specific calculation formula is: ; In the formula: is the decay rate; Iteratively train the multi-task time series weighted loss function until a preset threshold is reached or the maximum number of iterations is reached to obtain a trained efficiency decay trend prediction model.

[0037] Cross-channel fusion layer: Coupled analysis of the static features of the feature clustering channel and the dynamic time series features of the time series prediction channel is realized through an attention mechanism. Among them, the attention coupling function is: ; ; In the formula: is the attention mechanism weight; is the time series expression form of the actual operating efficiency of the water pump; is the cumulative operating time of the impeller, that is, the time series variable; is the time-varying extension of the time series data of the fault type q; is the predicted value of the actual operating efficiency of the water pump obtained after the coupled analysis of the static features and the dynamic time series features; is the time series value of the impeller wear coefficient; is the real-time power factor time series value; is the current waveform distortion rate time series value.

[0038] Based on the results of fault diagnosis and efficiency evaluation of the machine learning model, the visualization tracking of the pump efficiency evolution trend is realized through the digital twin visualization platform, and a decision support report containing optimization suggestions and maintenance strategies is generated.

[0039] Embodiment 2: This embodiment provides a system for monitoring the operating state of a pump unit based on digital twin technology, including: A digital twin model creation module that constructs a digital twin model of the pump unit, and the digital twin model integrates the equipment parameters and environmental sensing parameters of the pump unit; A data acquisition module that obtains the operating state data of the pump unit in real time through Internet of Things sensors, and the pump operating state data includes dynamic pressure values, instantaneous flow rates, power quality parameters, and fluid temperature parameters; An efficiency calculation module that dynamically calculates the actual operating efficiency of the pump unit based on the digital twin model and the real-time operating state data of the pump unit, and performs multi-dimensional correction; An evaluation and diagnosis module that combines the historical operating state data of the pump unit and the corrected actual operating efficiency, and performs fault diagnosis and efficiency evaluation of the pump unit through a machine learning model to obtain a fault diagnosis and evaluation report of the pump unit.

[0040] In the embodiments of the present application, "at least one" means one or more, and "a plurality" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent the situation where A exists alone, A and B exist simultaneously, or B exists alone. Where A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects before and after. "At least one of the following" and its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0041] Those of ordinary skill in the art can realize that the various units and algorithm steps described in the embodiments disclosed herein can be implemented by a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0042] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0043] In several embodiments provided in the present application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (hereinafter referred to as ROM), random access memory (hereinafter referred to as RAM), magnetic disks, or optical discs that can store program codes.

[0044] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for monitoring the operating status of a water pump unit based on digital twin technology, characterized in that: The following steps are involved: Constructing a digital twin model of the water pump unit, wherein the digital twin model integrates equipment parameters and environmental sensor parameters of the water pump unit; The operation status data of the water pump unit is obtained in real time through the Internet of Things sensor, and the water pump operation status data includes dynamic pressure value, instantaneous flow rate, power quality parameters and fluid temperature parameters; Based on the digital twin model and the real-time operating status data of the water pump unit, the actual operating efficiency of the water pump unit is dynamically calculated, and multi-dimensional correction is performed; Combined with the historical operating status data of the water pump unit and the corrected actual operating efficiency, the fault diagnosis and efficiency evaluation of the water pump unit are carried out through the machine learning model to obtain the fault diagnosis and evaluation report of the water pump unit.

2. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 1 is characterized in that: The equipment parameters include: a water pump geometric characteristic parameter group and a metering device characteristic parameter group, wherein: The pump geometric characteristic parameter group includes the inner diameter of the pump inlet and outlet water pipes and the motor baseline efficiency; The characteristic parameter group of the metering device includes the pulse constant of the watt-hour meter, the transformation ratio of the voltage transformer, and the installation elevation of the pressure sensor.

3. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 1 is characterized in that: The power quality parameters specifically include: current waveform distortion rate, characteristic harmonic component amplitude spectrum and fundamental power factor.

4. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 1 is characterized in that: The multi-dimensional correction includes applying nozzle structure parameters to correct fluid flow rate, using temperature compensation algorithm and mechanical wear factor to dynamically correct head, and introducing power quality parameters to dynamically calibrate the actual operating efficiency of the water pump.

5. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 4 is characterized in that: The fluid flow rate is corrected using the nozzle structure parameters, specifically: ； ； In the formula, is the water inlet flow rate; is the outlet flow rate; Q is the flow rate; is the inner diameter of the water inlet pipe; is the inner diameter of the outlet pipe.

6. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 5 is characterized in that: The temperature compensation algorithm and mechanical wear factor are used to dynamically correct the lift, specifically: ； in: ； In the formula, H is the head; is the pump outlet pressure; is the water pump inlet pressure; is the liquid density under standard working water temperature; g is the acceleration of gravity; T is the real-time water temperature; is the standard working water temperature; is the thermal expansion coefficient of the pipeline; is the impeller wear coefficient; It is the elevation of the water pump outlet pressure gauge; It is the elevation of the water pump inlet pressure gauge; The accumulated running time of the impeller; is the impeller wear acceleration threshold; is the impeller decay rate.

7. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 4 is characterized in that: The power quality parameters are introduced to dynamically calibrate the actual operating efficiency of the water pump and calculate the actual operating efficiency of the water pump. The specific formula is: ； In the formula, is the actual operating efficiency of the pump; is the overall efficiency of the pump; is the motor baseline efficiency; is the real-time power factor; is the rated power factor; is the harmonic loss coefficient; THD is the current waveform distortion rate.

8. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 1, characterized in that: The machine learning model adopts a dual-channel heterogeneous architecture, including: Feature clustering channel: Generate equipment health status clustering center based on improved weighted K-means algorithm; the objective function of the improved weighted K-means algorithm is: ； Where: J is the objective function of the improved weighted K-means algorithm; C is the total number of clusters; l is the cluster index of the cluster; i is the pump sample index; is the time decay function; is the actual operating efficiency of the i-th water pump; is the actual operating efficiency cluster center of the lth cluster; is the production duration balance coefficient; is the commissioning time of the i-th water pump; is the cluster center of the production time of the lth cluster; m is the total number of fault types; is the fault type index, representing the jth fault type and the qth fault type respectively; is the fault correlation matrix element; is the indicator value of whether the qth fault type occurs in the i-th water pump; is the average occurrence frequency of the qth fault in the lth cluster; Time series prediction channel: The LSTM network is used to build an efficiency decay trend prediction model; the LSTM network objective function is to design a multi-task time series weighted loss function based on the LSTM cross entropy loss function, specifically: ； In the formula, It is a multi-task temporal weighted loss function; is the classification cross entropy loss; is the signal-to-noise ratio weight adjustment coefficient; is the phase compensation coefficient; is the dynamic attenuation weight; is the predicted efficiency value in time segment t; is the actual efficiency value in time segment t; is the wear leveling factor; Calculate the KL divergence; is the real-time probability distribution of the d-th feature cluster; is the baseline probability distribution of the dth feature cluster; D is the number of feature clusters; d is the index of the feature cluster; is the time window length; Cross-channel fusion layer: The attention mechanism is used to realize the coupling analysis of the static features of the feature clustering channel and the dynamic timing features of the timing prediction channel.

9. The method for monitoring the operating status of a water pump unit based on digital twin technology according to claim 8 is characterized in that: The specific calculation formula of the fault correlation matrix element is: ； Where: is the historical number of times fault type j and fault type q occur simultaneously; is the total number of times fault type j occurs alone; is the total number of times fault type q occurs alone; To prevent the denominator from being a constant of 0; .

10. A system for monitoring water pump efficiency and faults based on digital twin technology, characterized in that: include: A digital twin model creation module is used to build a digital twin model of a water pump unit, wherein the digital twin model integrates equipment parameters and environmental sensor parameters of the water pump unit; The data acquisition module acquires the operating status data of the water pump unit in real time through the Internet of Things sensor, and the water pump operating status data includes dynamic pressure value, instantaneous flow rate, power quality parameters and fluid temperature parameters; An efficiency calculation module, which dynamically calculates the actual operating efficiency of the water pump unit based on the digital twin model and the real-time operating status data of the water pump unit, and performs multi-dimensional correction; The evaluation and diagnosis module combines the historical operating status data of the water pump unit and the corrected actual operating efficiency, performs fault diagnosis and efficiency evaluation of the water pump unit through a machine learning model, and obtains a fault diagnosis and evaluation report of the water pump unit.

Images