A method for automatic verification of initial state of simulation model of digital twin system of nuclear power plant

Abstract

The present invention relates to the field of nuclear power simulation technology, and specifically to a method for automatically verifying the initial state of a simulation model of a digital twin system of a nuclear power plant. The method comprises the following steps: obtaining user requirements, and starting automatic verification according to the user requirements; calling the real-time numerical values ​​of model parameters of each subsystem in a preset steady-state simulation model and the corresponding unit historical data according to user requirements, and comparing and verifying the unit historical data and the real-time numerical data of model parameters for designated important parameters; and judging the operation status of the designated subsystem model under stable operating conditions or transient accident conditions. The beneficial effect is that the initialization module based on this method can automatically reach the actual steady-state operation data state of the power plant at a specified time in a relatively short time for the running simulation model, so that the digital twin system has the same capability as the state of the nuclear reactor system at a specified time.

Description

Technical Field

[0001] The present invention relates to the field of nuclear power simulation technology, and in particular to a method for automatically verifying the initial state of a simulation model of a digital twin system of a nuclear power plant. Background Art

[0002] At present, in the field of nuclear reactor operation, the operation of the unit mainly relies on the technical capabilities of the operators such as the main control room operators, and the handling of abnormal problems often needs to rely on personal experience and decision-making. Therefore, most of the operating events in nuclear power plants are related to human factors. On the other hand, with the long-term operation of the unit, there will be a large number of unit operation plan improvements and technical transformation needs, but the research and verification of nuclear power plant operation plans and technical transformation plans still lack technical means. At present, it mainly relies on single software calculations and identification tests of single equipment, or relies on expert reviews to make technical decisions. Therefore, it is necessary to study digital twin technology, and develop intelligent operation support tools through online intelligent detection, diagnosis and analysis technologies to solve the problems faced in the operation field.

[0003] To realize the operation digital twin of the nuclear reactor system, the digital twin system must have the same ability as the nuclear reactor system at a specified time. However, since the nuclear reactor system is a complex system, how to obtain key and sufficient data from its massive operation data and act on the simulation model so that the simulation model can automatically initialize to the same state as the nuclear reactor system is one of the difficulties. Therefore, usually an initialization method is designed, which can achieve a steady-state simulation model consistent with the power plant operation data based on the actual operation data of the nuclear power plant at a certain moment. At the same time, after the steady-state simulation model is constructed, it is necessary to fully carry out verification and confirmation work, such as comparing and verifying the model from various steady-state power steps, start-up and shutdown processes, accident processes, etc. The actual operation data of the reactor is verified and the reliability of the steady-state simulation model operation is evaluated. Summary of the invention

[0004] The purpose of the present invention is to provide a method for automatic verification of the initial state of a nuclear power plant digital twin system simulation model, so as to establish a steady-state simulation model consistent with the power plant operation data based on the actual operation data of the nuclear power plant at a certain moment, and at the same time realize automatic verification of whether the steady-state simulation model conforms to the actual operation conditions under various operating conditions, so as to prove and evaluate the suitability of the model for simulating the behavior of the nuclear power plant during assumed transient and accident periods.

[0005] The technical solution of the present invention is as follows: A method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model, comprising the following steps:

[0006] S1 obtains user requirements, and starts automatic verification according to the user requirements, wherein the user requirements include stable operating condition verification and transient accident condition verification, the stable operating condition verification is used to verify the specified subsystem model at a specified time under the stable operating condition, and the transient accident condition verification is used to verify the specified subsystem model at a specified time under the accident condition;

[0007] S2 calls the real-time values ​​of the model parameters of each subsystem in the preset steady-state simulation model and the corresponding historical data of the unit according to user needs, and compares and verifies the historical data of the unit and the real-time value data of the model parameters for the designated important parameters, and judges the operation status of the designated subsystem model under stable operation conditions or transient accident conditions through preliminary statistical analysis, wherein the real-time values ​​of the model parameters of each subsystem in the steady-state simulation model are obtained via a preset initialization module, and the initialization module is used to establish a steady-state simulation model consistent with the operation data of the power plant according to the actual operation data of the nuclear power plant at a certain moment, and the historical data of the unit is obtained via the actual unit;

[0008] S3 performs in-depth statistical analysis based on the real-time values ​​of the model parameters, the historical data of the unit and the calculation results of step S2 to further determine the operation status of the specified subsystem model under stable operating conditions or transient accident conditions.

[0009] In step S2, the preliminary statistical analysis includes obtaining the error comparison values ​​of the important parameters, and judging the operation status of the designated subsystem model under stable operation conditions or transient accident conditions based on the preliminary analysis of the error comparison values. If the error comparison value exceeds the preset error range, it is judged that the operation of the designated subsystem model is abnormal; if the error comparison value is within the preset error range, it is judged that the operation of the designated subsystem model is normal.

[0010] In step S3, the in-depth statistical analysis includes single parameter statistical analysis and multiple parameter statistical analysis;

[0011] The single parameter statistical analysis is used to perform statistical calculations on the unit historical data of a single parameter and the real-time numerical values ​​of the model parameters, and to perform data comparison, trend comparison, error curve drawing, error bar drawing, and the statistical calculation is a combination of one or more of the mean value, variance, slope, and confidence range interval;

[0012] The statistical analysis of the multiple parameters is used to obtain the influence curve of the important parameters along with the sensitive influencing parameters to show the change trend, wherein the sensitive influencing parameters are the parameters that affect the change of the important parameters.

[0013] In step S2, establishing a steady-state simulation model consistent with the power plant operation data according to the actual operation data of the nuclear power plant at a certain moment includes:

[0014] S21 data processing: processing the actual operation data of the unit to obtain tracking point data, and establishing unit data mapping between the tracking point data and the thermal software calculation data;

[0015] S22 performs matching calculations, obtains a reset IC based on the unit data mapping, resets the digital twin system with the IC, decouples the system after determining the model boundary parameter conditions of each subsystem model, and automatically starts calculations and parameter corrections based on the agreed boundary parameter conditions, iterating the calculations until the initialization calculations are completed.

[0016] The data processing step S21 includes the following steps:

[0017] S211 data management, performing one or more combinations of singular value removal, data completion, data smoothing, and data feature extraction on the actual operation data;

[0018] S212 data coordination, based on the principles of conservation of mass, conservation of energy, conservation of momentum, and compliance with physical properties, the actual operating data after data governance is corrected.

[0019] In step S211, the Raida criterion is used to remove singular values.

[0020] And / or, in step S211, the data completion includes: if it is determined that the actual operation data is empty within a certain time period, tracing back in the external database until the time when there is data is traced back, and completing according to the data of the last time traced back,

[0021] And / or, in the step S211, the data smoothing processing includes: performing direct shrinkage and stacking comparison on the actual operation data, and performing weighted smoothing calculation to smooth and balance the actual operation data.

[0022] In step S212, data coordination includes the following steps:

[0023] S2121 performs redundancy analysis on various measurement parameters according to the main constraints of the system, performs measurement redundancy analysis on the system, and determines a system or a corresponding measurement loop that can implement data coordination; if the system or the corresponding measurement loop has redundancy, it is determined that data coordination can be performed;

[0024] S2122 performs parameter correction on the system or measurement loop that is judged to be capable of data coordination according to the principles of conservation of mass, conservation of energy, and compliance with physical properties.

[0025] In step S22, the matching calculation includes the following steps:

[0026] S221 creates an initial state of a simulation model, which is used to determine the initial operating conditions of the digital twin system and fill in each simulation system model of the initial state according to the initial operating conditions;

[0027] S222 subsystem steady-state calculation is used to start each simulation system model that has been filled with the initial state and perform calculations respectively;

[0028] S223 joint steady-state calculation is used to store the results of the separate steady-state calculation in IC as a state recording point, and then gradually couple the simulation models of each subsystem, and compare the coupled system with the tracking data points to confirm whether the current model meets the technical requirements. If so, the current model is considered valid for the tracking data and the tracking is completed; if not, it falls back to the corresponding state save point in the state record for reconfirmation.

[0029] The step S221 of creating the initial state of the simulation model comprises the following steps:

[0030] S2211 obtains a reset IC and selects an IC with the most consistent parameter approximation with the tracking point data from the pre-stored IC data as the reset IC, and resets the digital twin system with the reset IC;

[0031] S2212 boundary parameter setting, for the reset digital twin system, adjusting the boundary parameters of each model in the digital twin system according to the tracking point data and the connection relationship between the systems to determine the boundary parameter conditions of each model;

[0032] S2213 decouples the model and decomposes the adjusted digital twin model system into independent computing processes.

[0033] The step S2211 includes:

[0034] S22111 pre-stores enough full range model ICs;

[0035] S22112 obtains and analyzes the core power parameters of the current reset data;

[0036] S22113 searches for the closest saved state from the model IC based on the core power parameter.

[0037] The steady-state calculation of the S222 subsystem includes the following steps:

[0038] S2221 subsystem operation confirmation is used to determine whether the initial state fitting of all simulation system model parameters is complete;

[0039] S2222 parameter tracking correction is used to correct non-critical parameters and key parameters of each simulation system.

[0040] In the step S2221, the subsystem operation is confirmed;

[0041] S22221 determines whether the initial state of the model is complete, and detects whether all non-zero values ​​of the simulation model have values. If so, compare with the preset reasonable interval to determine whether the non-zero value of the corresponding model is within the reasonable interval. If it is within the reasonable interval, proceed to step S22222. If it is not within the reasonable interval, an external reminder is issued.

[0042] S22222 Model decoupling confirmation is used to confirm whether all simulation models are in a decoupling state. If the decoupling is not complete or the decoupling fails, a message prompt will pop up and an error will be reported;

[0043] S22223 model operation startup and status monitoring: when it is determined that the model decoupling is completed, the simulation model operation is started, and the operation status of the model is monitored during the operation to ensure that the model operation is within the normal range and no abnormal deviation occurs. If a model abnormality occurs, the abnormal simulation model will be frozen and a message will pop up to prompt confirmation.

[0044] The step S231 joint adjustment steady-state calculation comprises the following steps:

[0045] S2311 state record is used to store the results of the separate regulation steady-state calculation in IC, record the current separate regulation state, and determine whether it is necessary to revert to the separate regulation stage according to the system operation state. When it is determined that the coupling process is abnormal or the parameter confirmation process fails, it reverts to the state point recorded in this process, and confirms whether it is necessary to further revert to the separate regulation stage;

[0046] S2312 subsystem coupling is used to gradually couple the simulation models of each subsystem, including opening the interface identifiers between the simulation models of each subsystem one by one, so that the interface program enters the online interactive state;

[0047] S2313 steady-state tracking parameter confirmation, obtain the main parameters of each model after operation, and compare them with the tracking data points to confirm whether the current model meets the technical requirements. If it meets the technical requirements, the current model is considered to be valid for the tracking data and the tracking is completed; if it does not meet the requirements, it will fall back to the corresponding state save point in the state record for reconfirmation.

[0048] The beneficial effect of the present invention is that the initialization module based on this method can automatically bring the running simulation model to the actual steady-state operation data state of the power plant at a specified time in a relatively short time, so that the digital twin system has the same capabilities as the nuclear reactor system at the specified time. At the same time, the simulation model produced by the initialization module is automatically compared, verified and evaluated for user needs, and automatically verified to prove and evaluate the suitability of the model to simulate the behavior of the nuclear power plant during the assumed transient and accident periods. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a workflow diagram of the automatic verification method;

[0050] Figure 2 Schematic diagram of the workflow for generating a simulation model for the initialization module;

[0051] Figure 3 Provides a summary flow of the tracking points;

[0052] Figure 4 It is the initial state analysis process;

[0053] Figure 5 Retrieve logical processes for IC;

[0054] Figure 6 It is the steady-state calculation process of the subsystem;

[0055] Figure 7 Run processes for subsystems;

[0056] Figure 8 It is the process of joint adjustment steady-state calculation;

[0057] Fig. 9 For online real-time comparison and verification. DETAILED DESCRIPTION

[0058] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0059] like Figure 1 As shown, the present invention discloses a method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model, comprising:

[0060] S1 obtains user requirements, and starts automatic verification according to the user requirements, wherein the user requirements include stable operating condition verification and transient accident condition verification, the stable operating condition verification is used to verify the specified subsystem model at a specified time under the stable operating condition, and the transient accident condition verification is used to verify the specified subsystem model at a specified time under the accident condition;

[0061] S2 calls the real-time values ​​of the model parameters of each subsystem in the preset steady-state simulation model and the corresponding historical data of the unit according to user needs, and compares and verifies the historical data of the unit and the real-time value data of the model parameters for the designated important parameters, and judges the operation status of the designated subsystem model under stable operation conditions or transient accident conditions through preliminary statistical analysis, wherein the real-time values ​​of the model parameters of each subsystem in the steady-state simulation model are obtained via a preset initialization module, and the initialization module is used to establish a steady-state simulation model consistent with the operation data of the power plant according to the actual operation data of the nuclear power plant at a certain moment, and the historical data of the unit is obtained via the actual unit;

[0062] S3 performs in-depth statistical analysis based on the real-time values ​​of the model parameters, the historical data of the unit and the calculation results of step S2 to further determine the operation status of the specified subsystem model under stable operating conditions or transient accident conditions.

[0063] The present invention provides a method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model, which is used to automatically verify whether the steady-state simulation model obtained by an initialization module conforms to actual operating conditions under various working conditions.

[0064] Among them, as a preferred solution, Figure 2 As shown, in this embodiment, the initialization module in step S2, that is, establishing a steady-state simulation model consistent with the power plant operation data according to the actual operation data of the nuclear power plant at a certain moment, includes the following steps:

[0065] S21 data processing: processing the actual operation data of the unit to obtain tracking point data, and establishing unit data mapping between the tracking point data and the thermal software calculation data;

[0066] S22 matching calculation, obtaining a reset IC according to the unit data mapping, resetting the digital twin system with the IC, performing system decoupling after determining the model boundary parameter conditions of each subsystem model, and automatically starting calculation and parameter correction with the agreed boundary parameter conditions, iterating the calculation until the initialization calculation is completed;

[0067] During the steady-state operation of the power plant, the measured steady-state data has a small vibration due to the disturbance of multiple complex factors. In the digital twin system, since the simulation model simplifies the disturbance factors, the actual simulation model steady-state cannot be calculated through the theoretical model to obtain data that is consistent with the fluctuating state of the power plant. The simulation model is also unable to identify steady-state information based on the fluctuating data source. Therefore, the data values ​​of the continuous power plant data segments required for the tracking point calculation should be noise-reduced and self-consistent.

[0068] In step S21, as a preferred solution, data governance and data coordination are performed on the actual operation data to obtain available tracking point data, wherein data governance is used to manage the quality of the actual operation data; the data coordination process mainly focuses on detecting parameter mismatches that may be introduced by measurement errors of sensors, and calculating and verifying the parameters, and specifically includes the following steps:

[0069] S211 data management, performing one or more combinations of singular value removal, data completion, data smoothing, and data feature extraction on the actual operation data;

[0070] S212 Data coordination: correct the actual operation data after data governance according to the principles of conservation of mass, conservation of energy, conservation of momentum, and physical property compliance;

[0071] Furthermore, in step S211, the Raida criterion is used to remove singular values, and singular values ​​of the actual operating data are removed, that is, values ​​with deviations greater than 3σ are removed as singular values.

[0072] And / or, in step S211, data completion includes: if it is determined that the actual operation data is empty within a certain time period, tracing back in the external database until the time when there is data is traced back, and completing according to the data of the last time traced back to fill in the missing values ​​of the actual operation data.

[0073] And / or, in step S211, the data smoothing process includes: performing direct shrinkage and stacking comparison on the actual operation data, and performing weighted smoothing calculations to smooth and balance the actual operation data (especially sensor data) to reduce the impact of accidental data mutations. For the actual operation data obtained, since the sampling samples are large, more than the number required for data coordination, and the data is subject to various interferences over a period of time. In order to condense the amount of data information itself, and not to ignore the useful information that the original measurement point sampling can bring, it is necessary to perform various smoothing processes before the data coordination calculation.

[0074] In step S212, data coordination is mainly to solve the problem of correcting data that does not conform to physical mechanisms such as conservation of mass and conservation of energy generated during the operation of the nuclear power plant, so that the corrected parameters can conform to the physical mechanism equations. In this embodiment, data coordination includes the following steps:

[0075] S2121 performs redundancy analysis on various measurement parameters according to the main constraints of the system, performs measurement redundancy analysis on the system, and determines the system or corresponding measurement loop that can implement data coordination; if the system or the corresponding measurement loop has redundancy, it is judged that data coordination can be performed. The higher the redundancy, the higher the accuracy of the coordination calculation. The constraints in the data coordination problem of the thermal system mainly include: mass balance conditions, energy balance conditions, momentum balance conditions, physical property equations, etc.

[0076] S2122 performs parameter correction on the system or measurement loop that is judged to be capable of data coordination according to the principles of conservation of mass, conservation of energy, and compliance with physical properties.

[0077] In this embodiment, the thermal parameters of the secondary circuit can be corrected based on the reliability of the thermal system sensors and other measurement parameters of the secondary circuit. Usually, the measurement parameters with low reliability are calibrated based on the measurement parameters with the highest reliability. It includes two parts: basic calibration and thermal balance calibration. The basic calibration completes pressure calibration, temperature calibration, liquid level calibration and flow calibration. The thermal balance calibration completes the secondary energy balance calibration of the generator, the efficiency calibration of the turbine stage group and the thermal balance calibration of the condenser.

[0078] In the initialization step, after the unit data mapping is completed, the process goes to step S22, where the unit effective data is assigned to the corresponding model parameters, and the new initial condition state is stored based on the platform initial IC (initial condition). After the model calculation is imported, the deviation is judged, and the parameters with large deviations are automatically adjusted, and the iterative calculation is performed until the initialization calculation is completed.

[0079] In this embodiment, as a preferred solution, step S22 matching calculation includes the following steps:

[0080] S221 creates an initial state of the simulation model to determine the initial operating conditions of the digital twin system. Further, Figure 3 As shown, step S221 creates the initial state of the simulation model including:

[0081] S2211 obtains a reset IC and selects an IC with the most consistent parameter approximation with the tracking point data from the pre-stored IC data as the reset IC, and resets the digital twin system with the reset IC;

[0082] S2212 boundary parameter setting, for the digital twin system after reset, the boundary parameters of each model in the digital twin system are adjusted according to the tracking point data and the connection relationship between the systems to determine the boundary parameter conditions of each model.

[0083] S2213 decouples the model and decomposes the adjusted digital twin model system into independent computing processes.

[0084] In this embodiment, as a preferred solution, Figure 4 As shown, in step S2211, obtaining the reset IC includes the following steps:

[0085] S22111 pre-stores enough full-range model ICs. Since the online initialization function considered in this project only includes online initialization during power operation, the pre-stored IC saves 51 IC states from full power to zero power reduction process at 2% power intervals and 20 IC states from zero power to 40% full power, a total of 71 pre-stored IC state points. According to the relationship between electric power Pe, average temperature Tavg of a circuit and IC number N, a search function relationship is established:

[0086] S22112 obtains and analyzes the core power parameters of the current reset data. In this embodiment, the current core power data can be obtained from the shared memory.

[0087] S22113 searches for the closest saved state from the model IC based on the core power parameter;

[0088] Furthermore, in this embodiment, if Figure 5As shown, searching for the closest saved state includes the following steps: comparing the core power with the rated power to calculate the normalized power. If the normalized power is greater than 40%, searching for the IC closest to the power from full power downwards for reset; if the normalized power is less than 40%, first determine whether the process is a power increase or power decrease based on power and temperature, and then search for the closest IC based on power after determining the process.

[0089] In step S2212, the boundary parameter setting of each simulation system is jointly determined based on the tracking point data and the connection relationship between the systems. Each boundary parameter should have its own setting method. For each simulation system, it is only necessary to design its own output boundary parameters. The following lists the overall boundary parameters of the step system, and each system performs boundary parameter setting on the following parameters: 1) The boundary parameter setting of the main coolant system mainly includes the setting of the interface boundary with the core system, chemical volume system, residual drainage system, injection system, containment system, main feed water system, auxiliary feed water system, evaporator sewage system, main steam system and other systems; 2) The core system mainly includes the parameter setting of the boundary with the main coolant system; 3) The flow network system includes a single-phase flow network simulation model, a two-phase flow network simulation model and an equipment model. According to the difference in the output interface parameters, the boundary of the flow network system is divided into a pressure flow boundary and a thermal boundary, which correspondingly includes the pressure flow boundary output parameter setting and the thermal boundary output parameter setting. 4) The main boundaries of the power generation system include: the mechanical power transmitted from the turbine to the generator side through the main shaft (turbine control and regulation system), the turbine lubrication top shaft and the turning system; the initial voltage, frequency and phase angle of the equivalent generator (balance node type) of the external power grid; the initial active power and reactive power of the equivalent load of the external power grid.

[0090] Among them, for the secondary circuit main steam-water system, it mainly includes the boundary setting of the steam system, condenser, condensate system, deaerator and feed water system: a) Steam system boundary: The steam system is initialized as a whole with the pipe network system from the steam generator outlet to the low-pressure cylinder last stage blade outlet. Therefore, the steam system boundary parameter setting is mainly to determine the boundary with the steam generator and the boundary with the condenser. b) The condenser is a device that connects the two-phase flow network and the single-phase flow network system. The boundary parameter setting of the condenser is mainly to determine the boundary between the condenser and the steam system and the boundary with the condensate system; c) The condensate system is a pipe network system from the condensate outlet of the condenser to the inlet of the deaerator. Therefore, the boundary parameter setting of the condensate system is mainly to determine the boundary with the deaerator and the boundary with the condenser; d) The deaerator is a device that connects the condensate system and the feed water system. The boundary parameter setting of the deaerator is mainly to determine the boundary between the deaerator and the condensate system and the boundary with the feed water system. e) The feedwater system is a pipe network system from the outlet of the deaerator to the feedwater inlet of the steam generator. Therefore, the boundary parameter setting of the feedwater system is mainly to determine the boundary with the deaerator and the boundary with the steam generator.

[0091] In the digital twin model system, the connection identifier between systems is composed of a series of logical variables. The connection identifier is named in the format of system name_system name, where the former system name is the current simulation system, and the latter system name is the connected simulation system. For example, in the main coolant system, the connection identifier with the core system is named RCP_CRS. When the connection identifier is in the false state, it means that the parameters between the simulation systems are not transmitted to each other, and the interface parameter values ​​are the set boundary values. When the connection identifier is in the true state, the boundary parameters between the simulation systems begin to be transmitted to each other. Therefore, in step S2213, when it is judged that the connection identifier is in the false state, the boundary parameter interaction code between the simulation systems will be skipped, and the parameters will not be transmitted.

[0092] like Figure 6 As shown, step S222 of the subsystem steady-state calculation is used to start each simulation system model that has been filled with the initial state, and perform calculations respectively, including:

[0093] S2221 subsystem operation confirmation is used to determine whether the initial state fitting of all simulation system model parameters is complete;

[0094] S2222 non-key parameter tracking and correction, which directly maps the actual power plant operation data to the model data of non-key parameters, and corrects the real-time calculation results of non-key parameters of each simulation system;

[0095] S2223 Model operation startup and status monitoring, when it is determined that the model decoupling is completed, the simulation model operation is started. And the operation status of the model is monitored during the operation to ensure that the model operation is within the normal range and no abnormal deviation occurs. If a model abnormality occurs, the abnormal simulation model will be frozen and a message will pop up to prompt confirmation. During the steady-state calculation of the subsystem, the system will execute three processes: subsystem operation, parameter tracking correction (coarse adjustment), and key parameter tracking correction (fine adjustment). The coarse adjustment and fine adjustment processes involve the back-substitution process of the joint adjustment stage. Each process in this stage must wait for all simulation model systems to be completed before entering the next process.

[0096] In step S2221, if each model meets the start-up conditions, all subsystem simulation models are loaded into the simulator platform and started to run. In this embodiment, further, as Figure 7 As shown, the judgment of whether the initial state fitting of the simulation system model parameters is complete includes the following steps:

[0097] S22211 determines whether the initial state of the model is complete, and detects whether all non-zero values ​​of the simulation model have values. If so, compare them with the preset reasonable interval to determine whether the non-zero value of the corresponding model is within the reasonable interval. If it is within the reasonable interval, proceed to step S2212. If it is not within the reasonable interval, the user needs to be prompted to analyze whether there are any abnormalities in the coordination and expansion process of the simulation data.

[0098] S22212 Model decoupling confirmation is used to confirm whether all simulation models are in a decoupling state. If the decoupling is not complete or the decoupling fails, a message will pop up and an error will be reported;

[0099] In this embodiment, step S22211 performs detection of gas phase temperature, liquid phase temperature, pressure, and cavitation fraction for the node type, and performs detection of liquid phase flow rate and gas phase flow rate for the connecting pipe type. When the non-zero values ​​of the corresponding models at each position of each system meet the requirements, it is judged that the initial states of each type are complete.

[0100] The following uses temperature detection as an example to explain the initial state of the model. First, based on the temperature and load curves of each component in the system during operation, the temperature range at each position is estimated as the corresponding reasonable interval (as shown in Table 1 below); then the temperature value of the corresponding position in the initial operating condition of the simulation model is obtained and compared with the reasonable interval. If the temperature value is within the reasonable interval, the initial state of the model is judged to be complete.

[0101] Table 1 Temperature and load curve

[0102] Location unit Temperature lower limit Temperature limit Core exit ℃ 291.4 329.8 Average temperature ℃ 291.4 310.0 core entrance ℃ 291.4 292.4 Steam generator steam temperature ℃ 283.6 291.4

[0103] In step S2222, for the main coolant system model, the adjustment order of each equipment parameter is as follows: pressure vessel → hot pipe section → stabilizer → pressure relief box → evaporator → transition section pump and cold pipe section; in the core system model, the axis power distribution, control rod value and critical boron concentration parameters are corrected respectively; in the flow network system model: the flow network model correction process is carried out in the following order: (1) equipment status; (2) main process parameter error calculation and correction, including: a) pipeline conductance; b) turbine stage group efficiency curve; c) regulating valve characteristic curve and pipeline conductance; d) heat exchanger heat transfer coefficient; e) pump characteristic curve; in the power generation system model, for the main generator simulation model, including no-load curve characteristic distribution and generator additional loss correction; for the external power grid load model, including the correction of the external power grid load active initial value and the external power grid load reactive initial value; for the excitation system model, including the correction of the excitation resistance of the excitation machine. The model correction of the steam turbine control system includes parameter correction of three types of control algorithms: 1. The output value of the RS trigger at that time, which affects the status of each valve, switch command and the number of steps of automatic control; 2. The output value of the counter, which affects the accuracy of time and logic; 3. The output value of the PID controller, which affects the opening status of the valve and controller.

[0104] S2223 key parameter tracking and correction is used to correct the key parameters of each simulation system to avoid long-term deviation from the expected target value during the calculation process of the key parameters.

[0105] In this embodiment, the key parameters include but are not limited to: the primary circuit average temperature, pressurizer water level, primary circuit flow and other key parameters of the reactor primary circuit, as well as the main steam flow, evaporator water level, main feed water flow and other key parameters of the secondary circuit. It is necessary to monitor the calculated values ​​of each key parameter in real time to prevent the calculation process from deviating from the expected target value for a long time. When the key parameter deviates from the expected target value for more than 10 seconds, it is necessary to automatically intervene and load the adjustment module to correct the key parameter. As a preferred solution, the correction method may include adjusting the flow rate using parameters such as energy loss coefficient and conductance, adjusting the temperature using temperature, internal energy or enthalpy value initialization, adjusting the pressure distribution using pressure, temperature or internal energy initialization, adjusting the liquid level using cavitation fraction and liquid level initialization, etc.

[0106] like Figure 8 As shown, step S223 of the joint adjustment steady-state calculation is used to couple and confirm the steady-state model achieved by the separate adjustments in steps until all simulation model systems are fully combined and reach the joint adjustment steady state. If it is determined that the interface does not match or becomes unstable after coupling, the unstable system will be decoupled and rolled back to the subsystem steady state for re-correction until the system completes the joint adjustment steady state. It is then determined that the model initialization process is completed and the tracking process ends.

[0107] As a preferred solution, in this embodiment, the joint regulation steady-state calculation includes the following steps:

[0108] S2231 status record is used to store the results of the separate adjustment steady-state calculation in IC, record the current separate adjustment state, and determine whether it is necessary to fall back to the separate adjustment stage according to the system operation state. After coupling, if the joint adjustment calculation formula shows that the calculation deviation of some systems in the multi-objective coupling adjustment optimization is too large or the calculation fails, it will fall back to the state point recorded in this process, and set the current optimization parameter value to a disabled value or the current optimization direction to an incorrect direction. If instantaneous oscillation or instantaneous calculation failure occurs after coupling, it is necessary to further fall back to the separate adjustment stage to redefine the boundary parameters between subsystems and adjust the parameter distribution within the subsystem.

[0109] S2233 steady-state tracking parameter confirmation obtains the main parameters of each model after operation, and compares them with the tracking data points to confirm whether the current model meets the technical requirements. If it meets the technical requirements, the current model is considered to be valid for the tracking data and the tracking is completed; if it does not meet the requirements, it will fall back to the corresponding state save point in the state record for reconfirmation.

[0110] In step S2231, after the state recording process is completed in the separate adjustment steady-state calculation stage, the result of the separate adjustment steady-state calculation is stored in IC before the joint adjustment steady-state calculation process starts, and the current separate adjustment state is recorded to backup the pre-start data for the subsequent subsystem simulation model coupling. In this embodiment, the process will continue to use the IC save function of the simulation platform, and set the logic of whether to fall back to the separate adjustment stage. When an abnormality occurs in the subsequent coupling process or the parameter confirmation process fails, it will fall back to the state point recorded in this process, and the logic of this stage will confirm whether to further fall back to the separate adjustment stage.

[0111] In step S2232, subsystem coupling is the process of gradually coupling the simulation models of each subsystem after the state recording process. In this process, the interface identifiers between the simulation models of each subsystem will be opened one by one, so that the interface program enters the online interactive state. Furthermore, in this embodiment, the coupling sequence of the subsystem models is main coolant system-core physical system-main steam system-condensate system-main feed water system-electrical system-control system. The continuous operation time of each system coupling needs to be greater than 1 minute, and ensure that the deviation of boundary parameters and main operating parameters within 1 minute is not greater than 0.5%.

[0112] In step S2232, after the subsystem simulation model coupling is completed, that is, after a period of stable operation, the main parameters of the model are compared with the tracking data points to confirm whether the current model meets the technical requirements. If the technical requirements are met, the current model is considered to be valid for the tracking data, that is, the tracking is completed. If the requirements are not met, it will fall back to the state save point for reconfirmation.

[0113] After the matching calculation, it is necessary to compare and verify the model obtained from the matching calculation from multiple aspects such as various steady-state power steps, start-up and shutdown processes, and accident processes to determine whether the model obtained from the matching calculation is reasonable.

[0114] In step S1, user requirements include: comparative analysis of steady-state operation and normal operating conditions (such as 100% FP steady-state, 75% FP steady-state, 50% FP steady-state, reactor start and shutdown), comparative analysis of transient accident conditions (such as turbine shutdown, LOCA, SGTR, MSLB, SBO accidents and other accidents in accident procedures and safety analysis reports).

[0115] In step S2, the comparison between the verification module and the actual unit is mainly carried out by comparing the measured values ​​of each sensor with the calculated values ​​of the model, mainly the parameters of the gas-liquid phase, temperature, pressure, flow rate, flow velocity, internal energy, cavitation fraction, enthalpy, dryness and other important fluid mechanics and heat transfer parameters.

[0116] Preferably, Fig. 9As shown, the preliminary statistical analysis includes obtaining the error comparison values ​​of the important parameters, and judging the operation of the designated subsystem model under the stable operation condition or the transient accident condition based on the preliminary analysis of the error comparison values. If the error comparison value exceeds the preset error range, it is judged that the operation of the designated subsystem model is abnormal. If the error comparison value is within the preset error range, it is judged that the operation of the designated subsystem model is normal.

[0117] Furthermore, the main operations of the comparison and verification module include: 1) obtaining the following information based on user needs: a) selecting comparison conditions; b) selecting the corresponding unit data time period or unit sample to obtain a comparison reference group for each condition; c) determining the comparison parameters and parameter fluctuation range; d) determining the comparison error range; 2) then performing automatic testing and online debugging comparison of the conditions, performing a visual data comparison of the unit's historical data and the real-time values ​​of the model parameters for important parameters, issuing an over-error alarm, and drawing error lines; 3) performing error analysis on real-time data, i.e., calculating and displaying the data obtained during the calculation process; 4) the data is stored in the database, and the error results can be traced back.

[0118] Preferably, in step S3, the in-depth statistical analysis includes single parameter statistical analysis and multiple parameter statistical analysis; the single parameter statistical analysis is used to perform statistical calculations on the unit historical data of a single parameter and the real-time numerical value of the model parameter, and perform data comparison, trend comparison, error curve drawing, error bar drawing, and the statistical calculation is a combination of one or more of the mean value, variance, slope, and confidence range interval;

[0119] The statistical analysis of multiple parameters is used to obtain the influence curve of important parameters along with sensitive influencing parameters to show the change trend, wherein the sensitive influencing parameters are parameters that affect the change of the important parameters;

[0120] For a single parameter, there are two data sources, one is the unit data, and the other is the real-time calculation data. For each data source, statistical calculations such as mean, variance, slope, and confidence interval are required. For these two data sources, data comparison and trend comparison are performed, error curves are calculated, and error bar plots are drawn;

[0121] For multiple parameters, it is generally the relationship between important parameters and sensitive influencing parameters. Obtain the influence curve of important parameters with input parameters to show the change trend. Generate the influence curve with input parameters as X-axis and output important parameters as Y-axis. For the situation where multiple parameters affect the same output, perform sensitivity analysis.

[0122] In addition, the result analysis can be used to generate a verification report based on the graphical data conclusions. By selecting the output parameters and output content, a comparative verification report for multiple output parameters can be generated.

Claims

1. A method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model, characterized in that: The steps include: S1 obtains user requirements, and starts automatic verification according to the user requirements, wherein the user requirements include stable operating condition verification and transient accident condition verification, the stable operating condition verification is used to verify the specified subsystem model at a specified time under the stable operating condition, and the transient accident condition verification is used to verify the specified subsystem model at a specified time under the accident condition; S2 calls the real-time values ​​of the model parameters of each subsystem in the preset steady-state simulation model and the corresponding historical data of the unit according to user needs, and compares and verifies the historical data of the unit and the real-time value data of the model parameters for the designated important parameters, and judges the operation status of the designated subsystem model under stable operation conditions or transient accident conditions through preliminary statistical analysis, wherein the real-time values ​​of the model parameters of each subsystem in the steady-state simulation model are obtained via a preset initialization module, and the initialization module is used to establish a steady-state simulation model consistent with the operation data of the power plant according to the actual operation data of the nuclear power plant at a certain moment, and the historical data of the unit is obtained via the actual unit; In step S2, establishing a steady-state simulation model consistent with the power plant operation data according to the actual operation data of the nuclear power plant at a certain moment includes: S21 data processing: processing the actual operation data of the unit to obtain tracking point data, and establishing unit data mapping between the tracking point data and the thermal software calculation data; S22 matching calculation, obtaining a reset IC according to the unit data mapping, resetting the digital twin system with the IC, performing system decoupling after determining the model boundary parameter conditions of each subsystem model, and automatically starting calculation and parameter correction with the agreed boundary parameter conditions, iterating the calculation until the initialization calculation is completed; In step S22, the matching calculation includes the following steps: S221 creates an initial state of a simulation model, which is used to determine the initial operating conditions of the digital twin system and fill in each simulation system model of the initial state according to the initial operating conditions; S222 subsystem steady-state calculation is used to start each simulation system model that has been filled with the initial state and perform calculations respectively; S223 Joint debugging steady-state calculation is used to store the results of the separate debugging steady-state calculation in IC as a state record point, and then gradually couple the simulation models of each subsystem, and compare the coupled system with the tracking data point to confirm whether the current model meets the technical requirements. If the technical requirements are met, the current model is considered to be valid for the tracking data and the tracking is completed; if the requirements are not met, it will fall back to the corresponding state save point in the state record for reconfirmation; S3 performs in-depth statistical analysis based on the real-time values ​​of the model parameters, the historical data of the unit and the calculation results of step S2 to further determine the operation status of the specified subsystem model under stable operating conditions or transient accident conditions.

2. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: In step S2, the preliminary statistical analysis includes obtaining the error comparison values ​​of the important parameters, and judging the operation status of the specified subsystem model under stable operation conditions or transient accident conditions based on the preliminary analysis of the error comparison values. If the error comparison value exceeds the preset error range, it is judged that the operation of the specified subsystem model is abnormal; if the error comparison value is within the preset error range, it is judged that the operation of the specified subsystem model is normal.

3. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: In step S3, the in-depth statistical analysis includes single parameter statistical analysis and multiple parameter statistical analysis; The single parameter statistical analysis is used to perform statistical calculations on the unit historical data of a single parameter and the real-time numerical values ​​of the model parameters, and to perform data comparison, trend comparison, error curve drawing, error bar drawing, and the statistical calculation is a combination of one or more of the mean value, variance, slope, and confidence range interval; The statistical analysis of the multiple parameters is used to obtain the influence curve of the important parameters along with the sensitive influencing parameters to show the change trend, wherein the sensitive influencing parameters are the parameters that affect the change of the important parameters.

4. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: The data processing step S21 includes the following steps: S211 data management, performing one or more combinations of singular value removal, data completion, data smoothing, and data feature extraction on the actual operation data; S212 data coordination, based on the principles of conservation of mass, conservation of energy, conservation of momentum, and compliance with physical properties, the actual operating data after data governance is corrected.

5. A method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 4, characterized in that: In step S211, the Raida criterion is used to remove singular values. And / or, in step S211, the data completion includes: if it is determined that the actual operation data is empty within a certain time period, tracing back in the external database until the time when there is data is traced back, and completing according to the data of the last time traced back, And / or, in the step S211, the data smoothing processing includes: performing direct shrinkage and stacking comparison on the actual operation data, and performing weighted smoothing calculation at the same time, so as to smooth and balance the actual operation data.

6. A method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 4, characterized in that: In step S212, data coordination includes the following steps: S2121 performs redundancy analysis on various measurement parameters according to the main constraints of the system, performs measurement redundancy analysis on the system, and determines a system or a corresponding measurement loop that can implement data coordination; if the system or the corresponding measurement loop has redundancy, it is determined that data coordination can be performed; S2122 performs parameter correction on the system or measurement loop that is judged to be capable of data coordination according to the principles of conservation of mass, conservation of energy, and compliance with physical properties.

7. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: The step S221 of creating the initial state of the simulation model comprises the following steps: S2211 obtains a reset IC and selects an IC with the most consistent parameter approximation with the tracking point data from the pre-stored IC data as the reset IC, and resets the digital twin system with the reset IC; S2212 boundary parameter setting, for the reset digital twin system, adjusting the boundary parameters of each model in the digital twin system according to the tracking point data and the connection relationship between the systems to determine the boundary parameter conditions of each model; S2213 decouples the model and decomposes the adjusted digital twin model system into independent computing processes.

8. A method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 7, characterized in that: The step S2211 includes: S22111 pre-stores enough full range model ICs; S22112 obtains and analyzes the core power parameters of the current reset data; S22113 searches for the closest saved state from the model IC based on the core power parameter.

9. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: The steady-state calculation of the S222 subsystem includes the following steps: S2221 subsystem operation confirmation is used to determine whether the initial state fitting of all simulation system model parameters is complete; S2222 parameter tracking correction is used to correct non-critical parameters and key parameters of each simulation system.

10. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 9, characterized in that: In the step S2221, the subsystem operation is confirmed; S22211 determines whether the initial state of the model is complete, and detects whether all non-zero values ​​of the simulation model have values. If so, compare with the preset reasonable interval to determine whether the non-zero value of the corresponding model is within the reasonable interval. If it is within the reasonable interval, proceed to step S22212. If it is not within the reasonable interval, an external reminder is issued. S22212 Model decoupling confirmation is used to confirm whether all simulation models are in a decoupling state. If the decoupling is not complete or fails, a message will pop up and an error will be reported; S22213 model operation startup and status monitoring: when it is determined that the model decoupling is completed, the simulation model operation is started, and the operation status of the model is monitored during the operation to ensure that the model operation is within the normal range and no abnormal deviation occurs. If a model abnormality occurs, the abnormal simulation model will be frozen and a message will pop up to prompt confirmation.

11. The method for automatically verifying the initial state of a nuclear power plant digital twin system simulation model according to claim 1, characterized in that: The step S223 of joint adjustment steady-state calculation comprises the following steps: S2231 state record, used to store the results of the separate regulation steady-state calculation in IC, record the current separate regulation state, and determine whether it is necessary to revert to the separate regulation stage according to the system operation state. When it is determined that the coupling process is abnormal or the parameter confirmation process fails, revert to the state point recorded in this process, and confirm whether to further revert to the separate regulation stage; S2232 subsystem coupling is used to gradually couple the simulation models of each subsystem, including opening the interface identifiers between the simulation models of each subsystem one by one, so that the interface program enters an online interactive state; S2233 Steady-state tracking parameter confirmation, obtains the main parameters of each model after the operation is completed, and compares them with the tracking data points to confirm whether the current model meets the technical requirements. If it meets the technical requirements, the current model is considered to be valid for the tracking data and the tracking is completed; if it does not meet the requirements, it will fall back to the corresponding state save point in the state record for reconfirmation.

Images