Edge calculation-oriented flow measurement method and device for condensate water supply system

Abstract

The invention discloses a condensate water supply system flow measurement method and device oriented to edge calculation, and the method comprises the following steps: S1, extracting multiple sets of pump set operation parameters under typical working conditions based on an industrial Internet platform historical database; s2, performing simulation modeling on the condensate water supply system, and constructing a nonlinear mapping relation between a pipeline resistance coefficient and flow by combining a pipeline geometric configuration, fluid physical property parameters and pump mechanical characteristics to obtain a pipeline resistance equation; s3, the pipeline resistance equation and the centrifugal pump lift flow characteristic equation are combined, an equation set is solved through a Newton-Raphson iteration method, and a theoretical calculation flow value is obtained; and S4, comparing the theoretical calculation flow value with the real-time measurement flow value, and dynamically adjusting model parameters to ensure that a model error is smaller than a set threshold value. The method has the effect of high flow measurement precision of the condensate water supply system.

Description

Technical Field

[0001] The present invention relates to the technical field of secondary circuit feed water flow measurement in nuclear power plants, and in particular to a condensate feed water system flow measurement method and device for edge computing. Background Art

[0002] The primary function of a nuclear power plant's secondary feedwater system is to supply water to the steam generators and maintain their liquid level. The steam generators, through heat exchange, utilize the heat generated by the reactor to heat the feedwater and generate steam to propel the turbine generators for power generation. Feedwater is delivered by the main feedwater pump and the starting feedwater pump. Excessive flow can cause the steam generator liquid level to be too high, affecting the quality of the steam produced. Excessive moisture in the steam can damage turbine blades. Too little flow can cause the steam generator liquid level to be too low, preventing the adequate removal of reactor heat. This wastes power and can also lead to the risk of core overheating.

[0003] Traditional flow measurement in nuclear power plant feedwater branches currently relies primarily on hardware sensors, such as electromagnetic flowmeters and differential pressure transmitters. These sensors present the following challenges: 1. Long-term operation of the sensors is susceptible to high temperatures and high pressures, resulting in reduced accuracy. 2. Due to the long operating cycles of nuclear power plants, sensor problems require downtime for maintenance, impacting overall operational efficiency and increasing maintenance costs. 3. Flow sensor installation locations are limited, requiring flowmeter installation to maintain a 10-degree straight pipe section and a 5-degree straight pipe section, resulting in insufficient effective measurement point coverage. 4. Flow measurement in water supply systems is significantly affected by operating conditions such as temperature and pressure, requiring specially designed flow sensors to withstand extreme operating conditions. Consequently, traditional flow sensors may not meet these requirements.

[0004] Some have tried using ultrasonic flow meters, but this is also limited by factors such as installation conditions. Others have tried to calculate flow using pump and pipeline characteristics. Most people set the pipeline resistance to a constant, which is a simplified form derived from the Darcy-Weisbach Equation in fluid mechanics:

[0005]

[0006] The above equation is only valid under ideal conditions, such as in nuclear power plant seawater circulation pipelines. It should be noted that the pipeline resistance KADM is not a universal constant; its value depends on whether the flow regime is laminar or turbulent, the fluid properties, and the pipeline geometry and conditions, such as roughness and valve opening. The traditional assumption of a static resistance coefficient does not reflect the actual pipeline resistance, resulting in low flow rate accuracy. Summary of the Invention

[0007] In view of this, it is necessary to provide a condensate water system flow measurement method and device for edge computing to effectively solve the technical problem of low flow measurement accuracy in the condensate water system.

[0008] The present invention provides a method for measuring the flow rate of a condensate water supply system for edge computing, comprising the following steps:

[0009] Step S1: extracting multiple sets of pump group operating parameters under typical working conditions based on the industrial Internet platform historical database;

[0010] Step S2: Simulate and model the condensate water supply system, and construct a nonlinear mapping relationship between the pipeline resistance coefficient and the flow rate by combining the pipeline geometry, fluid physical parameters, and pump mechanical characteristics to obtain the pipeline resistance equation;

[0011] Step S3: Combine the pipeline resistance equation and the centrifugal pump head flow characteristic equation, and solve the equations by Newton-Raphson iteration method to obtain a theoretical calculated flow value;

[0012] Step S4: Compare the theoretically calculated flow value with the real-time measured flow value, and dynamically adjust the model parameters to ensure that the model error is less than a set threshold.

[0013] Preferably, the step S1 is specifically as follows:

[0014] Real-time operating condition data is acquired and stored from the DCS system, and dynamic mapping between the real-time operating condition data and the simulation model is achieved through timestamp alignment. A multi-operating condition parameter set of the pump group is constructed based on the real-time operating condition data. The multi-operating condition data set is cleaned and the operating condition clustering is performed on the multi-operating condition data set to generate a standardized training data set.

[0015] Preferably, the step S2 is specifically as follows:

[0016] Using the BNL simulation platform, the thermodynamic boundary conditions and fluid physical parameters in historical data are integrated to establish a numerical simulation model of the condensate water supply system that includes pump mechanical characteristics, pipeline geometric configuration, and thermal-hydraulic coupling effects. The numerical simulation model is analyzed to obtain the pipeline resistance equation.

[0017] Preferably, the step S3 is specifically as follows:

[0018] Combining the pipeline resistance equation with the corresponding centrifugal pump head and flow characteristic equation, we get the nonlinear equation group:

[0019]

[0020] ;

[0021] in, is the centrifugal pump head, is the centrifugal pump flow rate, is the centrifugal pump head flow characteristic equation, is the pipe resistance coefficient, is the pipeline resistance equation, is the mass flow rate, is the density, To obtain the pipeline pressure drop between two pressure measuring points containing a centrifugal pump;

[0022] Obtain the total pressure difference of the pump pipeline system and set the initial flow value. The centrifugal pump includes a condensate pump and a feed water pump. Perform two-dimensional interpolation on the head flow characteristic equation of the condensate pump through the parametric equation method, and perform one-dimensional interpolation on the pipeline resistance equation and the head flow characteristic equation of the feed water pump to obtain the theoretical calculated flow value.

[0023] Preferably, the step S4 is specifically as follows:

[0024] Obtain the real-time flow rate value measured by the sensor, calculate the difference between the theoretical calculated flow rate value and the real-time measured flow rate value, analyze the model accuracy and evaluate the model credibility based on the difference, and dynamically adjust the model parameters of the condensate feed water system numerical simulation model based on the model accuracy and model credibility to keep the model error less than the set threshold, and record the error data at the same time.

[0025] Preferably, it also includes:

[0026] The theoretically calculated flow value is fed back to the nuclear power plant industrial Internet platform, and fault pattern matching is performed based on the theoretically calculated flow value to identify system anomalies, locate faults, and achieve abnormal diagnosis.

[0027] Preferably, it also includes:

[0028] An embedded real-time computing engine is built to implement data interaction with the DCS system through the OPC UA protocol. Flow soft measurement results, model confidence, and abnormal diagnosis signals are integrated into the human-computer interaction interface to achieve embedded monitoring of soft measurements.

[0029] Preferably, it also includes:

[0030] Based on the deviation analysis between the theoretically calculated flow value and the real-time measured flow value, the frequency conversion adjustment parameters of the centrifugal pump are adjusted to maintain the optimal thermal efficiency.

[0031] The present invention also provides a condensate water supply system flow measurement device for edge computing, including a memory and a processor, wherein a computer program is stored on the memory, and when the computer program is executed by the processor, the condensate water supply system flow measurement method for edge computing is implemented.

[0032] Compared with the existing technology, the present invention has the following beneficial effects: the present invention proposes a soft-measurement method for water feed flow based on dynamic coupling modeling of centrifugal pump characteristics and pipeline resistance. Unlike the traditional static resistance coefficient assumption, the present invention collects multi-dimensional operating parameters of the pump group, combines the centrifugal pump's full-operating characteristic curve with pipeline resistance simulation modeling, and constructs a simulation model to solve the flow rate of the secondary circuit condensate feed water. Through a dynamic resistance compensation mechanism and multi-parameter fusion modeling, it achieves accurate reconstruction of flow parameters, controls the flow soft-measurement error within a small range, and meets the refined control requirements of the secondary circuit water feed system. In the event of failure of traditional flow meters in nuclear power systems, the present invention can still achieve continuous and stable monitoring of water feed flow. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0034] FIG1 is a flow chart of an embodiment of a method for measuring flow of a condensate feed water system for edge computing provided by the present invention;

[0035] Figure 2a yes Figure 1 A schematic structural diagram of an embodiment of a condensate pump of a condensate feed water system in the illustrated embodiment;

[0036] Figure 2b yes Figure 1 A schematic structural diagram of an embodiment of a feed water pump of a condensate feed water system in the illustrated embodiment;

[0037] Figure 3 yes Figure 1 An algorithm flow chart of an embodiment of a soft sensor algorithm in the illustrated embodiment;

[0038] Figure 4 yes Figure 1 A schematic diagram of an embodiment of a simulation platform in the illustrated embodiment;

[0039] Figure 5 yes Figure 1 A fitting curve diagram of an embodiment of the pipeline resistance coefficient in the embodiment shown;

[0040] Figure 6a yes Figure 1 A comparison diagram of the theoretically calculated flow rate value and the real-time measured flow rate value in the embodiment shown;

[0041] Figure 6b yes Figure 1 A graph showing an embodiment of measurement error values ​​in the illustrated embodiment;

[0042] Figure 7 yes Figure 1A schematic diagram of a platform embedded in an embodiment of a monitoring platform is shown. DETAILED DESCRIPTION

[0043] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0044] Example 1

[0045] See also Figure 1 In this embodiment, a condensate water system flow measurement method for edge computing includes the following steps:

[0046] Step S1: extracting multiple sets of pump group operating parameters under typical working conditions based on the industrial Internet platform historical database;

[0047] Step S2: Simulate and model the condensate water supply system, and construct a nonlinear mapping relationship between the pipeline resistance coefficient and the flow rate by combining the pipeline geometry, fluid physical parameters, and pump mechanical characteristics to obtain the pipeline resistance equation;

[0048] Step S3: Combine the pipeline resistance equation and the centrifugal pump head flow characteristic equation, and solve the equations by Newton-Raphson iteration method to obtain a theoretical calculated flow value;

[0049] Step S4: Compare the theoretically calculated flow value with the real-time measured flow value, and dynamically adjust the model parameters to ensure that the model error is less than a set threshold.

[0050] This embodiment addresses the problem of flow calculation errors caused by ignoring the dynamic changes of pipeline resistance in the prior art, and provides an online soft measurement method and dynamic optimization method based on fitting the pipeline resistance coefficient based on historical data and combining pump characteristics and fluid parameters.

[0051] Specifically, the condensate pump system structure of the nuclear power plant in this embodiment includes sensor arrangement, such as Figure 2a As shown: Condensate flows out of the condenser (CND), enters multiple condensate pumps through branched pipes, is transported to the shaft seal heater for heating, and then sent to the deaerator. When the unit load is low, the condensate recirculation valve will be opened to return some condensate to the condenser hot well to ensure that the flow rate of the condensate pump is not less than the minimum allowable value. Valves are installed on the pipeline to control the flow and direction of flow, and pressure gauges and thermometers are installed to monitor pressure and temperature to ensure the normal operation of the system. The structure of the nuclear power plant feed water pump system includes the sensor layout, such as Figure 2bAs shown: After condensate enters the deaerator for deoxygenation, it is pressurized by the feedwater pump to become feedwater, which is then heated again and delivered to the steam generator. When the unit load is low, the feedwater recirculation valve opens to return some feedwater to the deaerator, ensuring that the feedwater pump flow rate does not fall below the minimum allowable value. The feedwater system is equipped with a variety of monitoring and control devices for real-time monitoring of feedwater flow, pressure, temperature, and other parameters, and automatically adjusts system operation accordingly to ensure a stable supply and quality of feedwater.

[0052] against Figure 2a and Figure 2b For the condensate water supply system, first of all, real-time parameter collection is carried out: based on the industrial Internet platform, that is, the Data History Package historical database, multi-dimensional parameters such as the operating speed of the pump group, the inlet and outlet pressures, the medium temperature, etc. are collected, and multiple groups of typical working conditions are extracted, including the operating parameters of the pump group under transient processes such as starting and stopping the reactor and peak regulation, to provide a multi-dimensional data base for subsequent model construction. Secondly, system simulation modeling and: using the BNL simulation platform, the thermodynamic boundary conditions and fluid physical parameters in the historical data are integrated to establish a full-system numerical simulation model including the mechanical characteristics of the pump, the geometric configuration of the pipeline and the thermal-hydraulic coupling effect. Thirdly, resistance modeling: combining the full-condition characteristic curve of the centrifugal pump with the pipeline resistance simulation modeling, a mapping relationship between the dynamic pipeline resistance coefficient and the flow rate is established based on the historical data and the simulation model, and a dynamic resistance curve is constructed. . Again, flow solution: based on the centrifugal pump head-flow characteristic equation , simultaneous pipe resistance equations , solve the equations by Newton iteration method, and get the flow soft measurement value output theoretical calculation flow Finally, online verification: Comparison of calculated traffic With flow measurement , dynamically adjust model parameters to keep the error ≤ 2%, and record error data to ensure measurement traceability accuracy within the full range of working conditions.

[0053] The software algorithm flow of this embodiment is as follows Figure 3 As shown, the system reads measurement point data such as pressure, temperature, and speed from the real-time database of the Industrial Internet platform's time-series data. After preprocessing, the pump outlet pressure is used to determine whether the pump is operating. If not, the pump status is returned to the front-end display and stored in the Industrial Internet platform's time-series data history database. If it is operating, the measurement point data is input into the calculation model to determine whether the fluid state is laminar or turbulent, and the equation is solved using the Newton iteration method. Once the calculation results converge, they are output to the front-end display and stored in the Industrial Internet platform's time-series data history database.

[0054] This embodiment overcomes the limitations of traditional flow measurement devices in the complex operating conditions of nuclear power plant secondary condensate and feedwater systems. Through multi-parameter fusion modeling, it achieves precise reconstruction of flow parameters. This enables continuous and stable monitoring of feedwater flow in scenarios where traditional flowmeters fail, such as the high-radiation environment of the nuclear island, high-temperature and high-pressure media, and impurity accumulation in pipelines. This method not only replaces the measurement function of physical flow meters but also constructs a dynamic, adaptive digital twin model of the hydraulic system by collecting real-time data on pump current-head characteristic curves and pipe network pressure differential-flow characteristics.

[0055] Compared with the existing technology, this embodiment significantly improves the measurement accuracy: through the dynamic resistance compensation mechanism and multi-parameter coupling modeling, the flow soft measurement error can be controlled within ±2%, meeting the refined control requirements of the secondary water supply system; full life cycle cost optimization: reducing dependence on high-precision physical flow meters, hardware investment costs are reduced by more than 10%; support for online parameter updates and model self-correction, avoiding downtime maintenance due to sensor failure, and reducing annual maintenance costs by approximately 20%; intelligent operation and maintenance upgrade: the embedded monitoring platform integrates real-time flow prediction, equipment health assessment and abnormal warning functions, which can identify potential risks such as pipeline scaling and pump efficiency degradation in advance, realizing the transition from "passive maintenance" to "active maintenance".

[0056] This embodiment can replace physical flow meters: in complex operating conditions such as high temperature, high pressure, high radiation, or insufficient straight pipe length, it can replace traditional electromagnetic flowmeters, differential pressure transmitters, and other physical sensors to achieve non-invasive flow monitoring. It can also operate in parallel with existing flow meters to achieve redundancy verification: by comparing multi-source data, it improves measurement system reliability and provides an independent verification channel for nuclear safety-grade control systems.

[0057] Specifically, the step S1 is as follows:

[0058] Real-time operating condition data is acquired and stored from the DCS system, and dynamic mapping between the real-time operating condition data and the simulation model is achieved through timestamp alignment. A multi-operating condition parameter set of the pump group is constructed based on the real-time operating condition data. The multi-operating condition data set is cleaned and the operating condition clustering is performed on the multi-operating condition data set to generate a standardized training data set.

[0059] This embodiment adopts multi-source heterogeneous data fusion acquisition technology. The sensor data of the nuclear power system is collected and processed through the distributed control system DCS. The DCS is connected to various sensors through interfaces to collect temperature, pressure, flow, radiation level and other data in real time, and stores them in the industrial Internet platform database. The dynamic mapping of working condition data and simulation model is achieved through timestamp alignment. Based on the nuclear power plant industrial Internet platform industrial Internet platform historical database, a multi-working condition parameter set of the pump group covering transient processes such as starting and stopping the reactor and peak regulation is constructed. A standardized training data set is generated through data cleaning and working condition clustering. The pump group working condition parameters specifically include inlet pressure , outlet pressure , inlet medium temperature , outlet medium temperature , speed and measured flow .

[0060] Specifically, the step S2 is as follows:

[0061] The Bonuli simulation platform (BNL) is used to integrate thermodynamic boundary conditions and fluid physical parameters in historical data to establish a numerical simulation model of the condensate feed water system, which includes pump mechanical characteristics, pipeline geometric configuration, and thermal-hydraulic coupling effects. The numerical simulation model is analyzed to obtain the pipeline resistance equation.

[0062] This embodiment uses thermal-hydraulic coupling simulation modeling technology based on BNL's PowerBuilder thermal system modeling software to build a dynamic resistance coefficient-flow mapping model by combining the pipeline geometry, fluid physical parameters and pump mechanical characteristics. , achieving high-precision digital representation of the resistance characteristics of complex pipe network systems.

[0063] PowerBuilder thermal system modeling software is a simulation modeling tool widely used for generating units such as thermal power plants, wind farms, and photovoltaic power plants. PowerBuilder divides modules based on actual power plant thermal system equipment or equipment combinations, establishing conservation equations for mass, energy, and momentum. By analyzing the connections between modules, the software automatically identifies the fluid network topology implicit in the thermal system and automatically solves large fluid networks.

[0064] Figure 4 This is a schematic diagram of the BNL simulation platform. The nodes at both ends of the condensate pump calculation are the condenser pressure and the condensate main pressure, respectively. The condenser pressure corresponds to P319_N4 in the figure, and the condensate main pressure corresponds to P319_N9 in the figure. The three condensate pumps are located between the two nodes. The piping system between the two nodes can be simulated through a combination of pipes and valves.

[0065] Figure 5 The pipeline resistance coefficient curve is fitted, including laminar and turbulent lines. Combined with the condensate main pipe flow rate, condensate pump characteristics, and simulation verification, the piping system coefficient KADM is calculated. A curve showing the piping resistance coefficient KADM as it changes with flow rate is obtained, with laminar and turbulent lines. At low flow rates, such as during startup and shutdown, the pipeline is primarily laminar, while in steady state, turbulent flow is predominant.

[0066] Specifically, the step S3 is as follows:

[0067] Combining the pipeline resistance equation with the corresponding centrifugal pump head and flow characteristic equation, we get the nonlinear equation group:

[0068]

[0069] ;

[0070] in, is the centrifugal pump head, is the centrifugal pump flow rate, is the centrifugal pump head flow characteristic equation, is the pipe resistance coefficient, is the pipeline resistance equation, is the mass flow rate, is the density, To obtain the pipeline pressure drop between two pressure measuring points containing a centrifugal pump;

[0071] Obtain the total pressure difference of the pump pipeline system and set the initial flow value. The centrifugal pump includes a condensate pump and a feed water pump. Perform two-dimensional interpolation on the head flow characteristic equation of the condensate pump through the parametric equation method, and perform one-dimensional interpolation on the pipeline resistance equation and the head flow characteristic equation of the feed water pump to obtain the theoretical calculated flow value.

[0072] This embodiment uses the Newton-Raphson iteration method to solve the flow rate, and the convergence threshold is set to ±0.1%. In the nuclear power secondary circuit condensate feed water system, there are two sets of pump equipment, the condensate pump group and the feed water pump group. Generally, not all pumps are running. The pump outlet pressure and speed can be used to determine whether the pump is running. The characteristic equation of the running pump is and the corresponding pipe network system characteristic equation Simultaneously solve the nonlinear equations. Given the total pressure difference of the pump-pipe system, use the Newton-Raphson iterative method to solve the nonlinear equations. Set the initial guess flow value. When solving the variable frequency feed pump head flow characteristic equation, use the parametric equation method to perform two-dimensional interpolation. When solving the condensate pump head flow characteristic equation and the pipeline resistance equation, use one-dimensional interpolation. The pipeline resistance curve needs to calculate the Reynolds number using the initial guess flow value, and then determine whether to use the laminar flow line or the turbulent flow line, and finally calculate the real-time theoretical flow calculation value. .

[0073] Specifically, the step S4 is as follows:

[0074] Obtain the real-time flow rate value measured by the sensor, calculate the difference between the theoretical calculated flow rate value and the real-time measured flow rate value, analyze the model accuracy and evaluate the model credibility based on the difference, and dynamically adjust the model parameters of the condensate feed water system numerical simulation model based on the model accuracy and model credibility to keep the model error less than the set threshold, and record the error data at the same time.

[0075] By comparing the theoretically calculated flow value with the real-time measured flow value, the error is calculated and the model accuracy is analyzed. At the same time, statistical methods such as mean square error and standard deviation are used to evaluate the credibility of the model.

[0076] Specifically, it also includes:

[0077] The theoretically calculated flow value is fed back to the nuclear power plant industrial Internet platform, and fault pattern matching is performed based on the theoretically calculated flow value to identify system anomalies, locate faults, and achieve abnormal diagnosis.

[0078] After calculating the flow rate, this embodiment feeds the calculated flow rate back to the Industrial Internet platform, providing a data foundation for fault diagnosis and operational efficiency optimization, supporting subsequent diagnostic and optimization functions. By analyzing the deviation between the predicted and actual flow rates, system anomalies can be identified through fault pattern matching, quickly locating potential faults and implementing anomaly diagnosis. This helps operation and maintenance personnel identify equipment anomalies such as pump efficiency degradation, valve jams, or partial pipe blockages.

[0079] Specifically, it also includes:

[0080] An embedded real-time computing engine is built to implement data interaction with the DCS system through the OPC UA protocol. Flow soft measurement results, model confidence, and abnormal diagnosis signals are integrated into the human-computer interaction interface to achieve embedded monitoring of soft measurements.

[0081] This embodiment develops an embedded real-time computing engine based on flow calculation, implements millisecond-level data interaction with the power plant DCS system through the OPC UA protocol, integrates real-time flow soft measurement, and integrates flow soft measurement results, model confidence, and abnormal diagnosis signals into the human-machine interface (HMI). It supports multi-dimensional data visualization and provides multi-dimensional visual monitoring and closed-loop control interfaces through the HMI interface, forming an embedded intelligent monitoring platform.

[0082] Specifically, it also includes:

[0083] Based on the deviation analysis between the theoretically calculated flow value and the real-time measured flow value, the frequency conversion adjustment parameters of the centrifugal pump are adjusted to maintain the optimal thermal efficiency.

[0084] Based on the deviation analysis between the flow prediction value and the actual operating parameters, the frequency conversion adjustment parameters of the feedwater pump can be automatically optimized to ensure that the secondary circuit thermal system always maintains the optimal thermal efficiency, thus achieving refined energy-saving control while ensuring the safe operation of the nuclear power unit.

[0085] To verify this embodiment, the soft measurement method provided by this embodiment is compared with the flow measurement method in the prior art. Figure 6a and Figure 6b is a comparison diagram of soft measurement flow and measured flow, Figure 6aThe flow rate changes over time during the startup process. The dotted line is the value calculated by the soft measurement model, and the solid line is the value measured by the flow meter. The calculated value is 0 when the pump is not started. The error between the measured and calculated values ​​is small at other times. Figure 6b shown.

[0086] Figure 7 This is a schematic diagram of the embedded monitoring platform, showing the construction of the front-end large screen of the condensate water supply system calculation model, including the process Figure 1 The five sections are Flowchart 2, Control Valve, Feedwater Pump, and Condensate Pump. The feedwater pump section displays status detection of four physical quantities: head, flow, power, and efficiency; fault information history; and three sets of characteristic lines and operating points for the feedwater pump: flow-head, flow-power, and flow-efficiency.

[0087] Example 2

[0088] This embodiment provides a condensate water supply system flow measurement device for edge computing, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, it implements the condensate water supply system flow measurement method for edge computing described in Example 1.

[0089] The condensate water supply system flow measurement device for edge computing provided in this embodiment is used to implement the condensate water supply system flow measurement method for edge computing. Therefore, the technical effects of the condensate water supply system flow measurement method for edge computing are also possessed by the condensate water supply system flow measurement device for edge computing, which will not be repeated here.

[0090] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed in the present invention should be covered by the present invention.

Claims

1. A condensate water system flow measurement method for edge computing, characterized in that: The following steps are involved: Step S1: extracting multiple sets of pump group operating parameters under typical working conditions based on the industrial Internet platform historical database; Step S2: Simulate and model the condensate water supply system, and construct a nonlinear mapping relationship between the pipeline resistance coefficient and the flow rate by combining the pipeline geometry, fluid physical parameters, and pump mechanical characteristics to obtain the pipeline resistance equation; Step S3: Combine the pipeline resistance equation and the centrifugal pump head flow characteristic equation, and solve the equations by Newton-Raphson iteration method to obtain a theoretical calculated flow value; Step S4: Compare the theoretically calculated flow value with the real-time measured flow value, and dynamically adjust the model parameters to ensure that the model error is less than a set threshold.

2. The edge computing-oriented condensate water supply system flow measurement method according to claim 1 is characterized in that: The step S1 is specifically as follows: Real-time operating condition data is acquired and stored from the DCS system, and dynamic mapping between the real-time operating condition data and the simulation model is achieved through timestamp alignment. A multi-operating condition parameter set of the pump group is constructed based on the real-time operating condition data. The multi-operating condition data set is cleaned and the operating condition clustering is performed on the multi-operating condition data set to generate a standardized training data set.

3. The edge computing-oriented condensate water supply system flow measurement method according to claim 1 is characterized in that: The step S2 is specifically as follows: A simulation platform is used to integrate thermodynamic boundary conditions and fluid physical parameters in historical data to establish a numerical simulation model of the condensate water supply system that includes pump mechanical characteristics, pipeline geometric configuration, and thermal-hydraulic coupling effects; the numerical simulation model is analyzed to obtain the pipeline resistance equation.

4. The edge computing-oriented condensate water supply system flow measurement method according to claim 1 is characterized in that: The step S3 is specifically as follows: Combining the pipeline resistance equation with the corresponding centrifugal pump head and flow characteristic equation, we get the nonlinear equation group: ； in, is the centrifugal pump head, is the centrifugal pump flow rate, is the centrifugal pump head flow characteristic equation, is the pipe resistance coefficient, is the pipeline resistance equation, is the mass flow rate, is the density, To obtain the pipeline pressure drop between two pressure measuring points containing a centrifugal pump; Obtain the total pressure difference of the pump pipeline system and set the initial flow value. The centrifugal pump includes a condensate pump and a feed water pump. Perform two-dimensional interpolation on the head flow characteristic equation of the condensate pump through the parametric equation method, and perform one-dimensional interpolation on the pipeline resistance equation and the head flow characteristic equation of the feed water pump to obtain the theoretical calculated flow value.

5. The edge computing-oriented condensate water supply system flow measurement method according to claim 1 is characterized in that: The step S4 is specifically as follows: Obtain the real-time flow rate value measured by the sensor, calculate the difference between the theoretical calculated flow rate value and the real-time measured flow rate value, analyze the model accuracy and evaluate the model credibility based on the difference, and dynamically adjust the model parameters of the condensate feed water system numerical simulation model based on the model accuracy and model credibility to keep the model error less than the set threshold, and record the error data at the same time.

6. The edge computing-oriented condensate water system flow measurement method according to claim 1, characterized in that: Also includes: The theoretically calculated flow value is fed back to the nuclear power plant industrial Internet platform, and fault pattern matching is performed based on the theoretically calculated flow value to identify system anomalies, locate faults, and achieve abnormal diagnosis.

7. The edge computing-oriented condensate water supply system flow measurement method according to claim 1, characterized in that: Also includes: An embedded real-time computing engine is built to implement data interaction with the DCS system through the OPC UA protocol. Flow soft measurement results, model confidence, and abnormal diagnosis signals are integrated into the human-computer interaction interface to achieve embedded monitoring of soft measurements.

8. The edge computing-oriented condensate water supply system flow measurement method according to claim 1, characterized in that: Also includes: Based on the deviation analysis between the theoretically calculated flow value and the real-time measured flow value, the frequency conversion adjustment parameters of the centrifugal pump are adjusted to maintain the optimal thermal efficiency.

9. A condensate water system flow measurement device for edge computing, characterized in that: It includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, it implements the edge computing oriented condensate water supply system flow measurement method as described in any one of claims 1-8.

Images