Simulation calculation method for high-pressure heater in nuclear power unit under de-coupling condition

Abstract

The application specifically relates to a simulation calculation method for a high-pressure heater split condition of a nuclear power unit, and belongs to the field of performance calculation and simulation of a secondary loop thermodynamic system of a nuclear power unit, and comprises the following steps: a mechanism model for performance calculation of the secondary loop thermodynamic system of the unit is established according to design condition data and normal operation condition data of the unit; a support vector regression model for prediction of key operation parameters of the secondary loop thermodynamic system of the unit is established and trained according to the normal operation condition data and high-pressure heater split condition data of the unit; and the mechanism model and the support vector regression model are fused to perform performance calculation of the secondary loop thermodynamic system of the unit after the high-pressure heater is split. The mechanism model and the support vector regression model are fused, and in the case that actual high-pressure heater split condition data is lacking, the accurate and rapid prediction of key operation states of the nuclear power unit after sudden splitting of one high-pressure heater of the nuclear power unit under normal operation conditions is realized.

Description

Technical Field

[0001] This invention relates to the field of performance calculation and simulation technology of the secondary loop thermal system of nuclear power units, specifically to a simulation calculation method for the disconnection of high-pressure heaters in nuclear power units that integrates mechanistic models and support vector regression. Background Technology

[0002] The high-pressure heater system in a nuclear power plant unit utilizes extracted steam from the turbine to heat the feedwater entering the steam generator, thereby improving cycle thermal efficiency. The high-pressure heater system consists of two parallel rows of high-pressure heaters, with feedwater flowing through each row at 50% capacity during normal operation. If one row of high-pressure heaters fails and disconnects, the system can continue operating through the other row. However, disconnecting a high-pressure heater causes it to stop extracting steam, resulting in a large amount of extracted steam originally intended for feedwater heating remaining in the turbine and continuing to operate. This causes a sudden increase in turbine power, leading to drastic fluctuations in turbine load. Simultaneously, the disconnection causes a rapid drop in the feedwater temperature entering the steam generator, which in turn causes a drop in the reactor primary coolant temperature. Due to the negative temperature coefficient effect of the moderator, positive reactivity is introduced into the reactor, leading to a rapid increase in reactor nuclear power and thermal power, potentially exceeding limits and severely impacting the safe operation of the nuclear power plant. After a high-pressure heater disconnection, the primary objective for operators is to quickly and precisely reduce turbine power to prevent reactor nuclear power from exceeding limits. Therefore, it is necessary to conduct accurate simulation calculations of the changes in unit status when the high-pressure heater is disconnected, so as to guide the unit operators to quickly and accurately control the unit operating parameters when a high-pressure heater disconnection occurs, thereby ensuring that the unit can continue to operate safely and economically.

[0003] Currently, the performance calculation methods for the secondary loop of nuclear power units are mainly divided into three categories: mechanism model, data-driven model, and hybrid model.

[0004] Mechanistic models require the establishment of mathematical models that accurately reflect physical processes. This places high demands on the professional knowledge of modelers, and the modeling process is tedious and time-consuming. Furthermore, the secondary loop system of a nuclear power plant has a complex structure, making it difficult to achieve accurate modeling. More importantly, pure mechanistic models exhibit large calculation deviations under transient conditions.

[0005] Data-driven models rely on training with historical operational data to achieve predictive objectives. However, they lack physical interpretability and have high requirements for the quality and quantity of training data. Nuclear power plants, with their high safety requirements, experience infrequent high-pressure heater shutdowns, making it difficult to obtain substantial data on actual high-pressure heater shutdowns. This becomes a core bottleneck restricting the application of data-driven models. A single data-driven model is insufficient for accurate performance calculations.

[0006] In recent years, some studies have attempted to combine mechanistic models with data-driven models to balance physical interpretability and prediction accuracy. However, existing hybrid modeling methods still face the following technical challenges when applied to the transient condition of high-pressure heater disconnection: High-pressure heater de-energization is a small-sample, extreme operating condition. Existing hybrid models typically rely on normal operating data or limited simulation data for training, and the data-driven part is prone to producing predictions that violate physical laws when extrapolating under de-energization conditions. In actual nuclear power plants, high-pressure heater de-energization events are extremely rare, and available historical data is very limited. Existing hybrid models usually require a large amount of labeled data covering various operating conditions to achieve good generalization performance, and are prone to overfitting under small-sample conditions, making it difficult to meet the requirements of practical engineering applications.

[0007] Therefore, it is necessary to develop a hybrid modeling method that can balance physical interpretability and prediction accuracy under small sample conditions of high-pressure heater decomposition, and can effectively capture nonlinear transient processes. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a simulation calculation method for the high-pressure heater disconnection operation of a nuclear power unit. This method integrates mechanistic models and support vector regression, and in the absence of actual high-pressure heater disconnection operation data, it can accurately and quickly predict the key operating status of a nuclear power unit after a sudden disconnection of a high-pressure heater under normal operating conditions. This provides reliable guidance for nuclear power unit operators to operate the unit and ensures the safe and economical operation of the nuclear power unit.

[0009] Support vector regression (SVR) excels in few-shot learning, nonlinear fitting, and generalization performance. Hybrid modeling methods that combine mechanistic models with SVR leverage the advantages of both, offering both physical interpretability and adaptability to complex nonlinear relationships. This makes them highly suitable for solving transient operating condition prediction problems in nuclear power units, such as high-temperature heater decoupling, where data is scarce.

[0010] This hierarchical collaborative architecture, which uses support vector regression to predict key operating parameters of the secondary thermal system of nuclear power units and a mechanistic model for physical calculations of the secondary thermal system, effectively integrates the advantages of both mechanistic modeling and support vector regression modeling methods. It is used to accurately predict the operating status of nuclear power units after the high-pressure heater is disconnected, providing key decision support for nuclear power unit operators, helping to prevent safety accidents such as nuclear power exceeding limits, and ensuring the safe operation of nuclear power units.

[0011] To achieve the above objectives, the present invention provides the following technical solution: A simulation calculation method for the disconnection operation of high-pressure heaters in nuclear power units, which integrates mechanistic models and support vector regression models, includes the following steps: S1. Based on the unit's design operating condition data and normal operating condition data, establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; The aforementioned mechanism model is used to solve the performance indicators of the secondary loop thermal system of the unit through iterative calculation under the condition of input key operating parameters of the secondary loop thermal system of the unit; S2. Based on the normal operating condition data of the unit and the disconnection condition data of the high-pressure heater, establish and train a support vector regression model for predicting key operating parameters of the secondary loop thermal system of the unit. The support vector regression model is used to quickly predict the key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected, based on the current normal operating conditions data of the unit. S3. The mechanism model and the support vector regression model are fused to form a simulation calculation model for the high-pressure heater disconnection condition of the nuclear power unit, which is used to perform performance calculations of the secondary loop thermal system of the unit after the high-pressure heater disconnection.

[0012] Prioritizing this step, S1 establishes a mechanistic model for calculating the performance of the unit's secondary loop thermal system based on the unit's design operating condition data and normal operating condition data, including the following steps: Obtain the unit's design operating condition data, and based on the unit's heat balance diagram, obtain the unit's operating data at different power levels, and establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; Data on the normal operating conditions of the unit are acquired to debug and verify the aforementioned mechanism model.

[0013] As a priority, in S1, the secondary loop thermal system of the unit includes a steam generator, a steam turbine, a steam-water separator reheater, a condenser, a condensate pump, a feedwater pump, a deaerator, high-pressure heaters at various levels, and low-pressure heaters at various levels. The primary side of the steam generator is connected to the reactor coolant circuit, receiving heat from the reactor and heating the feedwater on its secondary side to form main steam to supply the turbine for operation; The flow path of the steam turbine includes a high-pressure cylinder composed of stages and a low-pressure cylinder composed of stages; the main steam outlet of the steam generator is connected to the steam inlet of the high-pressure cylinder through the main steam pipeline; The steam-water separator reheater is configured to use high-pressure cylinder extraction steam and main steam to perform multi-stage reheating of high-pressure cylinder exhaust steam. Its steam inlet is connected to the high-pressure cylinder exhaust port through a pipeline, and its steam outlet is connected to the low-pressure cylinder inlet through a pipeline. The steam inlet of the condenser is connected to the exhaust port of the low-pressure cylinder; the inlet of the condensate pump is connected to the condensate outlet of the condenser. The low-pressure heaters of each stage are connected in series along the condensate flow direction. The water-side inlet of the first-stage low-pressure heater along the condensate flow direction is connected to the outlet of the condensate pump, and the water-side outlet of the last-stage low-pressure heater along the condensate flow direction is connected to the water-side inlet of the deaerator. The inlet of the feedwater pump is connected to the water-side outlet of the deaerator; Each stage of high-pressure heater is connected in series along the feedwater flow direction. The water-side inlet of the first stage high-pressure heater along the feedwater flow direction is connected to the outlet of the feedwater pump, and the water-side outlet of the last stage high-pressure heater along the feedwater flow direction is connected to the feedwater inlet of the steam generator. The extraction ports of each stage of the high-pressure cylinder are connected to the steam side of the corresponding high-pressure heaters through extraction pipes; the extraction ports of each stage of the low-pressure cylinder are connected to the steam side of the deaerator and the corresponding low-pressure heaters through extraction pipes.

[0014] As a preferred step, in S1, the mechanistic model iteratively calculates and solves for the performance indicators of the unit's secondary loop thermal system, including the following steps: S101. Calculate the heat exchange of each stage of the low-pressure heater based on the condensate flow rate, condensate temperature, condensate pressure, and the inlet and outlet water temperatures of each stage of the low-pressure heater. Calculate the heat exchange capacity of the deaerator and the high-pressure heaters at each stage based on the water flow rate, water temperature, water pressure, and inlet and outlet water temperatures of the high-pressure heaters at each stage. S102. Determine the specific enthalpy and specific entropy of the main steam based on its pressure and dryness; set the design efficiency of each stage of the high-pressure cylinder as the initial efficiency for iteration. Based on the extraction pressure and temperature at each extraction point of the high-pressure cylinder, calculate the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder; based on the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder, and combined with the efficiency of each section of the high-pressure cylinder, calculate the extraction specific enthalpy of each section of the high-pressure cylinder and use it as the input parameter for the next stage. Calculate the steam extraction flow rate at each extraction point of the high-pressure cylinder based on the heat exchange capacity of the high-pressure heater. S103. Calculate the exhaust flow rate and dryness of the high-pressure cylinder based on the exhaust parameters of the high-pressure cylinder. Based on the exhaust flow rate and dryness of the high-pressure cylinder, flow balance and heat balance calculations are performed on the steam-water separator reheater to obtain the inlet steam flow rate and inlet steam specific enthalpy of the low-pressure cylinder. S104. Calculate the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder based on the extraction pressure and extraction temperature of each extraction point in the low-pressure cylinder. Based on the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder, and combined with the efficiency of each stage of the low-pressure cylinder, the extraction specific enthalpy of each stage of the low-pressure cylinder is calculated and used as the input parameter for the next stage of the low-pressure cylinder. Based on the heat exchange of the low-pressure heater obtained from S101, the steam extraction flow rate at each stage extraction point of the low-pressure cylinder is calculated. S105. Calculate the exhaust flow rate and exhaust enthalpy of the low-pressure cylinder; To calculate the power output of a steam turbine, the power output of the steam turbine is equal to the sum of the work done by each stage of the steam turbine multiplied by the mechanical efficiency and the generator efficiency. The deviation between the calculated and measured values ​​of the turbine's power generation is calculated. If the deviation is less than the preset convergence threshold, the heat balance calculation is performed on the condenser section, and the calculation results are output. Otherwise, the efficiency of each stage of the high-pressure cylinder and low-pressure cylinder is adjusted, and the iteration calculation is repeated in S102 until the deviation meets the convergence condition. S106. When the performance of the secondary loop thermal system of the unit is required under the condition of high-pressure heater disconnection, modify the steam-water flow in the mechanism model to make it consistent with the steam-water flow of the secondary loop thermal system of the unit under the condition of high-pressure heater disconnection; then calculate according to the process of S101-S105 to obtain the performance index of the secondary loop thermal system of the unit.

[0015] Prioritizing this step, in S2, a support vector regression model for predicting key operating parameters of the unit's secondary loop thermal system is established and trained, including the following steps: S201. Training Data Acquisition and Dataset Partitioning A simulation model was built using EBSILON, a professional simulation software verified by historical operating data of nuclear power units. The high-pressure heater disconnection operation of nuclear power units under different seawater temperatures, turbine exhaust pressures and electrical power was simulated, thereby generating a simulation dataset for training the support vector regression model. The simulation dataset includes input features and target variables. Input features include seawater temperature, turbine exhaust pressure, and electrical power. Target variables include main steam pressure after the high-pressure heater is disconnected, main steam flow rate, extraction pressure at each stage, water-side temperature of each stage heater, and exhaust pressure. The simulation model has been validated using historical operating data of the nuclear power unit, and the relative error between the predicted and measured values ​​is controlled within ±1%. The generated simulation dataset is divided into training and testing sets according to a certain ratio; S202, Data Preprocessing The Z-score standardization method is used to standardize the input features of the support vector regression model; the formula for the Z-score standardization method is: ; in, These are standardized input feature values. These are the original input feature values. The mean of the original input feature values ​​on the training set. It is the standard deviation of the original input feature values; After training, the support vector regression model outputs standardized predicted values. These standardized predicted values ​​are then de-standardized to obtain the physical predicted values, using the following formula: ; in, These are physical predictions. It is a standardized prediction value. It is the mean of the standardized forecasts. It is the standard deviation of the standardized predicted values; S203, Training of Support Vector Regression Model The radial basis function (RBF) kernel is chosen as the default kernel function for the support vector regression model; the RBF kernel function has the following form: ; in: It is a radial basis kernel function; There are two eigenvectors and The square of the Euclidean distance between them; These are kernel function hyperparameters used to control the range of influence of the samples; S203, Hyperparameter Optimization Bayesian optimization algorithm is used to optimize the penalty hyperparameter C and kernel function hyperparameter of the support vector regression model. Automatic optimization includes the following steps: Determine the objective function: Using the mean squared error of the support vector regression model on the validation set as the objective function, the goal is to find the hyperparameter combination (C) that minimizes the mean squared error. The expression for MSE is: ; in, The objective function value is represented by the mean squared error; N is the sample size. It is the predicted value of the i-th sample. It is the true value of the i-th sample; Constructing a surrogate model: A Gaussian process is used as a surrogate model to fit the hyperparameter combination (C). The probability distribution relationship between the objective function value (MSE) and the objective function value (MSE). Selection of acquisition function: The expected improvement function is adopted as the acquisition function to balance the exploration of unknown hyperparameter regions and the utilization of known hyperparameter regions that are better, and to guide the selection of the next set of hyperparameters to be evaluated. Iterative optimization: Repeat the iterative process of "selecting the next set of hyperparameters based on the acquisition function - training the support vector regression model and calculating the MSE - updating the Gaussian process surrogate model" until the preset number of iterations or convergence condition is reached, ultimately finding the optimal combination of hyperparameters (C0) that minimizes the mean squared error of the support vector regression model on the given dataset. ).

[0016] Preferably, in step S3, the fusion of the mechanistic model and the support vector regression model specifically includes the following steps: S301, Prediction of key operating parameters of the unit's secondary circuit thermal system after the high-pressure heater is disconnected. Real-time data on the unit’s current normal operating conditions is acquired, standardized according to the method in S202, and then input into the support vector regression model trained in S2. The support vector regression model will output standardized predicted values ​​of key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected; The standardized predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected are de-standardized by the support vector regression model to obtain the physical predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected. S302, Mechanism Model-Driven and State Calculation The physical prediction values ​​of key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected are used as input to the mechanism model established in S1. The mechanism model is driven to complete the iterative calculation of the performance of the entire secondary loop thermal system of the unit and output the performance index of the secondary loop thermal system of the unit after the high-pressure heater is disconnected.

[0017] Preferably, the input to the support vector regression model is the normal operating condition data of the unit, including seawater temperature, turbine exhaust pressure and electrical power; the output of the support vector regression model is the key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected, including main steam flow rate, main steam pressure, water-side inlet and outlet temperatures of each stage heater, extraction pressure and exhaust pressure of each stage heater.

[0018] Preferably, the performance indicators of the secondary loop thermal system of the unit after the high-pressure heater is disconnected include turbine power, reactor power change trend, main steam flow, parameters of each heater and condenser vacuum.

[0019] The present invention provides a computer device, including a memory and a processor. The memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the above-mentioned simulation calculation method for the disconnection of high-pressure heaters in nuclear power units.

[0020] To address the aforementioned technical problems, the present invention provides a computer-readable storage medium storing computer-readable instructions, which, when executed, implement the steps of the above-mentioned simulation calculation method for the disconnection of high-pressure heaters in nuclear power units.

[0021] Compared with the prior art, the present invention has the following beneficial technical effects: The data bottleneck problem has been solved: by using proven nuclear power simulation software to generate training data, the problem of lack of real high-pressure heater disconnection data caused by the high safety requirements of nuclear power units has been effectively overcome, opening up a feasible path for the application of machine learning methods in key nuclear power safety scenarios.

[0022] The model combines accuracy and interpretability: It employs a hierarchical architecture—using a support vector regression (SVR) model to predict key system parameters and a mechanistic model for system computation—to fully leverage the advantages of SVR in small sample sizes and nonlinear fitting, while retaining the clear physical meaning and transparent computational process of the mechanistic model. The SVR model accurately predicts the changes in local system parameters at the instant of resolution, while the mechanistic model extrapolates these changes across the entire system in accordance with physical laws. This overcomes the inherent limitations of a single model and achieves high-precision, high-reliability simulation computation.

[0023] High computational efficiency: The Bayesian optimization algorithm is used to automatically optimize the hyperparameters of the support vector regression model. Compared with traditional grid search or random search, it can find the optimal combination of hyperparameters in fewer evaluation times, which significantly reduces the computational cost and time of model training.

[0024] Strong guiding significance: This invention can provide nuclear power unit operators with an accurate preview of the changes in unit status under the condition of high-pressure heater disconnection accident, which helps operators to formulate and optimize operating strategies in advance, achieve rapid and accurate unit control, effectively prevent safety accidents such as nuclear power exceeding limits, and improve the safety and economy of nuclear power plant operation. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of the secondary loop thermal system structure of a unit according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating a simulation calculation method for the disconnection of a high-pressure heater in a nuclear power unit according to an embodiment of the present invention. Figure 3 This is a flowchart illustrating the performance calculation of the secondary loop thermal system of a nuclear power unit according to a mechanism model of an embodiment of the present invention. Figure 4 This is a flowchart illustrating the training and optimization process of a support vector regression model according to an embodiment of the present invention. Detailed Implementation

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application.

[0027] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0028] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and specific embodiments.

[0029] This embodiment provides a simulation calculation method for the disconnection of high-pressure heaters in nuclear power units, which integrates a mechanistic model and a support vector regression model. The high-pressure heating system in the unit consists of two parallel high-pressure heaters. When one high-pressure heater disconnects, the feedwater inlet and outlet valves of that high-pressure heater are closed, so that 75% of the feedwater flow passes through the normally operating high-pressure heater, and the remaining 25% of the feedwater flow flows directly to the feedwater header through a bypass.

[0030] The simulation calculation method for the high-pressure heater disconnection condition of nuclear power units, which integrates the mechanistic model and the support vector regression model, as described in this embodiment, includes the following steps: S1. Based on the unit's design operating condition data and normal operating condition data, establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; The aforementioned mechanism model is used to solve the performance indicators of the secondary loop thermal system of the unit through iterative calculation under the condition of input key operating parameters of the secondary loop thermal system of the unit; S2. Based on the normal operating condition data of the unit and the disconnection condition data of the high-pressure heater, establish and train a support vector regression model for predicting key operating parameters of the secondary loop thermal system of the unit. The support vector regression model is used to quickly predict the key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected, based on the current normal operating condition data of the unit. The key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected include main steam flow rate, main steam pressure, water-side inlet and outlet temperatures of each stage heater, extraction steam pressure and exhaust steam pressure of each stage heater. S3. The mechanism model and the support vector regression model are fused to form a simulation calculation model for the high-pressure heater disconnection condition of the nuclear power unit, which is used to perform performance calculations of the secondary loop thermal system of the unit after the high-pressure heater disconnection.

[0031] like Figure 1 As shown, in this embodiment, the secondary loop thermal system of the unit includes a steam generator, a steam turbine, a steam-water separator reheater, a condenser, a condensate pump, a feedwater pump, a deaerator, a three-stage high-pressure heater, and a three-stage low-pressure heater. The primary side of the steam generator is connected to the reactor coolant circuit, receiving heat from the reactor and heating the feedwater on its secondary side to form main steam to supply the turbine for operation; The steam turbine is configured as a single-shaft, four-cylinder, six-exhaust structure. Its flow passage includes a high-pressure cylinder consisting of the first to third stages and a low-pressure cylinder consisting of the fourth to eighth stages. The main steam outlet of the steam generator is connected to the steam inlet of the high-pressure cylinder through the main steam pipeline. The steam-water separator reheater is configured to perform two-stage reheating of the high-pressure cylinder exhaust steam by using high-pressure cylinder extraction steam and main steam. Its steam inlet is connected to the high-pressure cylinder exhaust port through a pipeline, and its steam outlet is connected to the low-pressure cylinder inlet through a pipeline. The steam inlet of the condenser is connected to the exhaust port of the low-pressure cylinder; the inlet of the condensate pump is connected to the condensate outlet of the condenser. The three-stage low-pressure heaters include a No. 3 low-pressure heater, a No. 2 low-pressure heater, and a No. 1 low-pressure heater connected in series along the condensate flow direction. The water-side inlet of the No. 3 low-pressure heater is connected to the outlet of the condensate pump, and the water-side outlet of the No. 1 low-pressure heater is connected to the water-side inlet of the deaerator. The inlet of the feedwater pump is connected to the water-side outlet of the deaerator; The three-stage high-pressure heater includes a No. 1 high-pressure heater, a No. 2 high-pressure heater, and a No. 3 high-pressure heater connected in series along the feedwater flow direction. The water-side inlet of the No. 1 high-pressure heater is connected to the outlet of the feedwater pump, and the water-side outlet of the No. 3 high-pressure heater is connected to the feedwater inlet of the steam generator. The extraction ports of each stage of the high-pressure cylinder are connected to the steam side of the corresponding three-stage high-pressure heater through extraction pipes; the extraction ports of each stage of the low-pressure cylinder are connected to the steam side of the deaerator and the corresponding three-stage low-pressure heater through extraction pipes. Low-pressure heater No. 1 and low-pressure heater No. 2 are both composed of three parallel columns A, B and C, while high-pressure heater No. 3 and low-pressure heater No. 3 are both composed of two parallel columns A and B.

[0032] When the secondary loop thermal system of the unit is operating, the main steam generated on the secondary side of the steam generator enters the high-pressure cylinder to expand and do work. In the high-pressure cylinder, some of the steam that has done work is extracted and supplied to the three-stage high-pressure heaters for heating the feedwater in stages; the exhaust steam from the high-pressure cylinder enters the steam-water separator reheater, where steam-water separation is performed first, and then two-stage reheating is carried out using the extracted steam from the high-pressure cylinder and the main steam to increase the steam superheat and temperature before it is sent to the low-pressure cylinder to continue expanding and doing work. In the low-pressure cylinder, steam with appropriate parameters is extracted from different stages and supplied to the deaerator and the three-stage low-pressure heaters for heating the condensate and deaerating; the exhaust steam from the low-pressure cylinder is finally discharged into the condenser and cooled and condensed into water by the circulating water.

[0033] On the condensate-feedwater side, the condensate in the condenser is pressurized by the condensate pump and flows sequentially through the No. 3, No. 2 and No. 1 low-pressure heaters to absorb heat from the regenerative extraction steam for staged heating, and then enters the deaerator to complete thermal deaeration. The deaerated feedwater is pressurized by the feedwater pump and flows sequentially through the three-stage high-pressure heaters, and is further heated to the specified feedwater temperature by the high-pressure cylinder extraction steam, and finally returns to the steam generator to absorb heat from the reactor, forming a continuous thermal cycle.

[0034] In this embodiment, S1, based on the unit's design operating condition data and normal operating condition data, a mechanism model for calculating the performance of the unit's secondary loop thermal system is established, including the following steps: Obtain the unit's design operating condition data, and based on the unit's heat balance diagram, obtain the unit's operating data at different power levels, and establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; Data on the normal operating conditions of the unit are acquired to debug and verify the aforementioned mechanism model.

[0035] In this embodiment, S1, the iterative calculation of the mechanism model to solve for the performance indicators of the unit's secondary loop thermal system includes the following steps: S101, Heat balance calculation for regenerator Calculate the heat exchange of each stage of the low-pressure heater based on the condensate flow rate, condensate temperature, condensate pressure, and the inlet and outlet water temperatures of each stage of the low-pressure heater. Calculate the heat exchange capacity of the deaerator and the high-pressure heaters at each stage based on the water flow rate, water temperature, water pressure, and inlet and outlet water temperatures of the high-pressure heaters at each stage. The calculations of the thermodynamic properties of water and water vapor are based on the internationally recognized IAPWS-IF97 standard; S102, Calculation of the thermal process of the high-pressure cylinder Based on the pressure and dryness of the main steam, determine the specific enthalpy and specific entropy of the main steam; set the design efficiency of each stage of the high-pressure cylinder as the initial efficiency for iteration; Based on the extraction pressure and temperature at each extraction point of the high-pressure cylinder, calculate the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder; based on the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder, and combined with the efficiency of each section of the high-pressure cylinder, calculate the extraction specific enthalpy of each section of the high-pressure cylinder and use it as the input parameter for the next stage. Calculate the steam extraction flow rate at each extraction point of the high-pressure cylinder based on the heat exchange capacity of the high-pressure heater. S103, Heat balance calculation for steam-water separator reheater Calculate the exhaust flow rate and dryness of the high-pressure cylinder based on the exhaust parameters of the high-pressure cylinder. Based on the exhaust flow rate and dryness of the high-pressure cylinder, flow balance and heat balance calculations are performed on the steam-water separator reheater to obtain the inlet steam flow rate and inlet steam specific enthalpy of the low-pressure cylinder. S104, Calculation of thermal process in low-pressure cylinder Calculate the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder based on the extraction pressure and extraction temperature of each extraction point in the low-pressure cylinder. Based on the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder, and combined with the efficiency of each stage of the low-pressure cylinder, the extraction specific enthalpy of each stage of the low-pressure cylinder is calculated and used as the input parameter for the next stage of the low-pressure cylinder. Based on the heat exchange of the low-pressure heater obtained from S101, the steam extraction flow rate at each stage extraction point of the low-pressure cylinder is calculated. S105. Condenser thermal balance and power verification calculation Calculate the exhaust flow rate and exhaust specific enthalpy of the low-pressure cylinder; To calculate the power output of a steam turbine, the power output of the steam turbine is equal to the sum of the work done by each stage of the steam turbine multiplied by the mechanical efficiency and the generator efficiency. The deviation between the calculated and measured values ​​of the turbine's power generation is calculated. If the deviation is less than the preset convergence threshold, the heat balance calculation is performed on the condenser section, and the calculation results are output. Otherwise, the efficiency of each stage of the high-pressure cylinder and low-pressure cylinder is adjusted, and the iteration calculation is repeated in S102 until the deviation meets the convergence condition. S106, Mechanism model adaptation for high-pressure heater disconnection condition When the performance of the secondary loop thermal system of the unit is required under the condition of high-pressure heater disconnection, the steam-water flow in the mechanism model is modified to be consistent with the steam-water flow of the secondary loop thermal system of the unit under the condition of high-pressure heater disconnection. For example, 75% of the feedwater flow flows through the normally operating high-pressure heater, and the remaining 25% of the feedwater flow flows directly to the feedwater header through a bypass. Then, the performance indicators of the secondary loop thermal system of the unit are obtained by calculating according to the S101-S105 process.

[0036] In this embodiment, step S2 involves establishing and training a support vector regression model for predicting key operating parameters of the unit's secondary loop thermal system, including the following steps: S201. Training Data Acquisition and Dataset Partitioning Ideally, model training should be conducted using actual high-pressure heater disconnection data from nuclear power units. However, due to the high safety requirements of nuclear power unit operation, there is a lack of sufficient actual high-pressure heater disconnection data from nuclear power units to support model training. To address the lack of actual high-pressure heater disconnection data for nuclear power units, this invention employs EBSILON, a professional simulation software validated by historical operating data of nuclear power units and capable of accurately simulating high-pressure heater disconnection conditions, to construct a simulation model. The model simulates high-pressure heater disconnection conditions under different seawater temperatures, turbine exhaust pressures, and electrical power, thereby generating a simulation dataset for training a support vector regression model. The accuracy and reliability of the simulation software are crucial prerequisites for the effectiveness of subsequent support vector regression model training; therefore, this invention rigorously designs the process for establishing and validating the simulation model, specifically including the following steps: Based on the actual structural parameters and heat balance diagram of the unit's secondary loop thermal system, a simulation model completely consistent with the unit's secondary loop thermal system was built in the professional simulation software EBSILON. Taking the unit's VWO operating condition as the design operating condition, by comparing the design values ​​and simulation values ​​of each key operating parameter in the design operating condition simulation model, the relative error between the design values ​​and simulation values ​​of each key operating parameter is required to be controlled within ±1%. To verify the accuracy of the above simulation model, historical operating data of the unit under normal operating conditions for one year was collected for simulation model verification. The collected historical operating data included main steam flow rate, main steam pressure, water-side inlet and outlet temperatures of each stage heater, extraction steam pressure and exhaust steam pressure under different power, seawater temperature and turbine exhaust steam pressure. Based on typical steady-state operating conditions from the unit's historical operating data, the measured boundary conditions, including electrical power, seawater temperature, and turbine exhaust pressure, are used as inputs to the simulation model for simulation calculations. The output values ​​of the simulation model are compared with the measured values, and the relative error between the output values ​​and the measured values ​​is required to be controlled within ±1%. The output of the simulation model includes main steam pressure, main steam flow rate, extraction steam pressure at each stage, water-side temperature of each stage heater, and exhaust pressure. After verifying the accuracy of the simulation model, the simulation calculation of the high-pressure heater disconnection condition is carried out using the verified simulation model, and a simulation dataset is generated for subsequent support vector regression model training. In the simulation model, set up a case where column A or column B of the high-pressure heaters is cut off, that is, simulate the scenario where one column of high-pressure heaters is cut off due to a fault, while the other column of high-pressure heaters is operating normally. The simulation model calculates the main steam pressure, main steam flow, extraction steam pressure of each stage, water side temperature of each stage heater and exhaust steam pressure after the high-pressure heaters are cut off. Through the above simulation, a total of 30 simulation data samples of high-pressure heater disconnection conditions were generated, consisting of different boundary condition combinations composed of seawater temperature, turbine exhaust pressure and electric power. Each sample contains input features and target variables. The input features include seawater temperature, turbine exhaust pressure and electric power. The target variables include the main steam pressure, main steam flow rate, extraction pressure of each stage, water-side temperature of each stage heater and exhaust pressure after high-pressure heater disconnection. These together constitute the simulation dataset for subsequent support vector regression model training. This simulation dataset effectively solves the bottleneck problem of insufficient data on high-pressure heater disconnection events in actual nuclear power plants, and provides sufficient training samples for subsequent prediction of high-pressure heater disconnection operating parameters of nuclear power units based on Bayesian optimized support vector regression models. The generated simulation dataset is divided into training and test sets according to a certain ratio, for example, 80% is allocated to the training set and the remaining 20% ​​to the test set; S202, Data Preprocessing Since the support vector regression model is sensitive to the scale of the input data, it is necessary to standardize the input features of the support vector regression model so that each input feature is within the same scale range. This invention employs the Z-score normalization method to standardize the input features of the support vector regression model, effectively improving the training effect and prediction accuracy of the support vector regression model. The Z-score normalization method is not strongly affected by outliers and can provide more stable results for the complex data of most nuclear power units. The formula for the Z-score standardization method is: ; in, These are standardized input feature values. These are the original input feature values. The mean of the original input feature values ​​on the training set. It is the standard deviation of the original input feature values; After standardization, all input feature values ​​will become a distribution with a mean of 0 and a standard deviation of 1. After training, the support vector regression model outputs standardized predicted values. The standardized predicted values ​​output by the support vector regression model need to be de-standardized to obtain the physical predicted values, using the following formula: ; in, These are physical predictions. It is a standardized prediction value. It is the mean of the standardized forecasts. It is the standard deviation of the standardized predicted values; S203, Training of Support Vector Regression Model This invention selects the radial basis function (RBF) kernel as the default kernel function for the support vector regression model. The RBF kernel can effectively handle data with nonlinear relationships and exhibits good regression performance. The form of the RBF kernel is: ; in: There are two eigenvectors and The square of the Euclidean distance between them; These are kernel function hyperparameters used to control the range of influence of the samples; The magnitude of the value determines the distribution of the sample after it is mapped to a high-dimensional space; It is the radial basis function kernel, whose physical meaning is: to measure the relationship between two eigenvectors x and y. The similarity between two samples; when two samples are very close in the original space. When the kernel function value approaches 0, it approaches 1, indicating high similarity. When two samples are far apart, the kernel function value rapidly decays to 0, indicating extremely low similarity. In this way, when predicting new samples, the support vector regression model will give greater weight to support vectors with high similarity, thereby achieving localized nonlinear fitting. In the support vector regression model, there are several key hyperparameters that need to be tuned to obtain the best model performance. Among them, the penalty hyperparameter C affects the model's tolerance to errors in the training data; the kernel function hyperparameter γ controls the range of influence of the samples, and the value of γ determines the distribution of the samples after mapping to the high-dimensional space. During the internal training process of the Support Vector Regression (SVR) model, given a set of hyperparameters (C, γ), the SVR model obtains the Lagrange multipliers by solving a convex quadratic programming problem. , and bias b: ; After training, for any input feature vector x, the predicted output of the support vector regression model is: ; in, and All are Lagrange multipliers. For bias, m is the number of training samples; only The corresponding training samples, i.e. support vectors, contribute to the prediction results, while the coefficients of the remaining samples are zero, thus realizing the sparse representation of the support vector regression model. S203, Hyperparameter Optimization To further optimize the performance of the support vector regression model, this invention employs a Bayesian optimization algorithm to optimize two key hyperparameters of the support vector regression model: the penalty hyperparameter C and the kernel function hyperparameter. Automatic optimization is performed to improve the predictive performance of the support vector regression model and obtain the optimal support vector regression model. The Bayesian optimization algorithm finds the optimal combination of hyperparameters that minimizes the regression error by efficiently searching the hyperparameter space. Compared with traditional grid search and random search methods, this invention can find the optimal combination of hyperparameters with fewer function evaluations, significantly reducing the computational cost. The penalty hyperparameter C and kernel function hyperparameter of the support vector regression model were optimized using the Bayesian optimization algorithm. Automatic optimization includes the following steps: Determine the objective function: Using the mean squared error of the support vector regression model on the validation set as the objective function, the goal is to find the hyperparameter combination (C) that minimizes the mean squared error. The expression for MSE is: ; in, is the objective function value, and is the mean squared error. The smaller the value, the better the fit of the support vector regression model; N is the number of samples. It is the predicted value of the i-th sample. It is the true value of the i-th sample; Constructing a surrogate model: A Gaussian process is used as a surrogate model to fit the hyperparameter combination (C). The probability distribution relationship between the objective function value (MSE) and the objective function value (MSE). Selection of acquisition function: The expected improvement function is adopted as the acquisition function to balance the exploration of unknown hyperparameter regions and the utilization of known hyperparameter regions that are better, and to guide the selection of the next set of hyperparameters to be evaluated. Iterative optimization: Repeat the iterative process of "selecting the next set of hyperparameters based on the acquisition function - training the support vector regression model and calculating the MSE - updating the Gaussian process surrogate model" until the preset number of iterations or convergence conditions are reached, ultimately finding the optimal combination of hyperparameters that minimizes the mean squared error of the support vector regression model on the given dataset. ).

[0037] In this embodiment, step S3, fusing the mechanistic model with the support vector regression model, specifically includes the following steps: S301, Prediction of key operating parameters of the unit's secondary circuit thermal system after the high-pressure heater is disconnected. Real-time data on the unit's current normal operating conditions, including seawater temperature, turbine exhaust pressure, and electrical power, are acquired, standardized according to the S202 method, and then input into the support vector regression model trained in S2. The support vector regression model will output standardized predicted values ​​of key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected, including main steam flow, main steam pressure, water-side inlet and outlet temperatures of each stage heater, and changes in extraction and exhaust steam pressures at each stage. The standardized predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected are de-standardized by the support vector regression model to obtain the physical predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected. S302, Mechanism Model-Driven and State Calculation The physical prediction values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected are used as inputs to the mechanism model established in S1. The mechanism model is driven to complete the performance iteration calculation of the entire unit's secondary loop thermal system according to the process from S101 to S106, and outputs the performance indicators of the unit's secondary loop thermal system after the high-pressure heater is disconnected. The performance indicators of the secondary loop thermal system of the unit after the high-pressure heater is disconnected include turbine power, reactor power change trend, main steam flow, parameters of each heater and condenser vacuum.

[0038] As an implementation of the above method, the present invention provides an embodiment of a computer device, which corresponds to the embodiment of the above method.

[0039] The computer device described in this embodiment includes a memory and a processor. The memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, it implements the steps of the above-described simulation calculation method for the disconnection of the high-pressure heater of a nuclear power unit.

[0040] As an implementation of the above method, the present invention provides an embodiment of a computer-readable storage medium, which corresponds to the embodiment of the above method.

[0041] The computer-readable storage medium described in this embodiment stores computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the simulation calculation method for the disconnection of the high-pressure heater of a nuclear power unit as described above.

[0042] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A simulation calculation method for the disconnection operation of a high-pressure heater in a nuclear power unit, characterized in that, Includes the following steps: S1. Based on the unit's design operating condition data and normal operating condition data, establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; The aforementioned mechanism model is used to solve the performance indicators of the secondary loop thermal system of the unit through iterative calculation under the condition of input key operating parameters of the secondary loop thermal system of the unit; S2. Based on the normal operating condition data of the unit and the disconnection condition data of the high-pressure heater, establish and train a support vector regression model for predicting key operating parameters of the secondary loop thermal system of the unit. The support vector regression model is used to quickly predict the key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected, based on the current normal operating conditions data of the unit. S3. The mechanism model and the support vector regression model are fused to form a simulation calculation model for the high-pressure heater disconnection condition of the nuclear power unit, which is used to perform performance calculations of the secondary loop thermal system of the unit after the high-pressure heater disconnection.

2. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 1, characterized in that, Based on the unit's design operating condition data and normal operating condition data, a mechanistic model for calculating the performance of the unit's secondary loop thermal system is established, including the following steps: Obtain the unit's design operating condition data, and based on the unit's heat balance diagram, obtain the unit's operating data at different power levels, and establish a mechanism model for calculating the performance of the unit's secondary loop thermal system; Data on the normal operating conditions of the unit are acquired to debug and verify the aforementioned mechanism model.

3. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 1, characterized in that, In S1, the mechanistic model iteratively calculates and solves for the performance indicators of the unit's secondary loop thermal system, including the following steps: S101. Calculate the heat exchange of each stage of the low-pressure heater based on the condensate flow rate, condensate temperature, condensate pressure, and the inlet and outlet water temperatures of each stage of the low-pressure heater. Calculate the heat exchange capacity of the deaerator and the high-pressure heaters at each stage based on the water flow rate, water temperature, water pressure, and inlet and outlet water temperatures of the high-pressure heaters at each stage. S102. Determine the specific enthalpy and specific entropy of the main steam based on its pressure and dryness; set the design efficiency of each stage of the high-pressure cylinder as the initial efficiency for iteration. Based on the extraction pressure and temperature at each extraction point of the high-pressure cylinder, calculate the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder; based on the extraction specific entropy and ideal extraction specific enthalpy at each extraction point of the high-pressure cylinder, and combined with the efficiency of each section of the high-pressure cylinder, calculate the extraction specific enthalpy of each section of the high-pressure cylinder and use it as the input parameter for the next stage. Calculate the steam extraction flow rate at each extraction point of the high-pressure cylinder based on the heat exchange capacity of the high-pressure heater. S103. Calculate the exhaust flow rate and dryness of the high-pressure cylinder based on the exhaust parameters of the high-pressure cylinder. Based on the exhaust flow rate and dryness of the high-pressure cylinder, flow balance and heat balance calculations are performed on the steam-water separator reheater to obtain the inlet steam flow rate and inlet steam specific enthalpy of the low-pressure cylinder. S104. Calculate the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder based on the extraction pressure and extraction temperature of each extraction point in the low-pressure cylinder. Based on the extraction specific entropy and ideal extraction specific enthalpy of each extraction point in the low-pressure cylinder, and combined with the efficiency of each stage of the low-pressure cylinder, the extraction specific enthalpy of each stage of the low-pressure cylinder is calculated and used as the input parameter for the next stage of the low-pressure cylinder. Based on the heat exchange of the low-pressure heater obtained from S101, the steam extraction flow rate at each stage extraction point of the low-pressure cylinder is calculated. S105. Calculate the exhaust flow rate and exhaust enthalpy of the low-pressure cylinder; To calculate the power output of a steam turbine, the power output of the steam turbine is equal to the sum of the work done by each stage of the steam turbine multiplied by the mechanical efficiency and the generator efficiency. The deviation between the calculated and measured values ​​of the turbine's power generation is calculated. If the deviation is less than the preset convergence threshold, the heat balance calculation is performed on the condenser section, and the calculation results are output. Otherwise, the efficiency of each stage of the high-pressure cylinder and low-pressure cylinder is adjusted, and the iteration calculation is repeated in S102 until the deviation meets the convergence condition. S106, Mechanism model adaptation for high-pressure heater disconnection condition When the performance of the secondary loop thermal system of the unit is required under the condition of high-pressure heater disconnection, the steam-water flow in the mechanism model is modified to be consistent with the steam-water flow of the secondary loop thermal system of the unit under the condition of high-pressure heater disconnection. Then, following the S101-S105 process, the performance indicators of the unit's secondary loop thermal system are obtained.

4. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 1, characterized in that, In S2, a support vector regression model for predicting key operating parameters of the unit's secondary loop thermal system is established and trained, including the following steps: S201. Training Data Acquisition and Dataset Partitioning A simulation model was built using EBSILON, a professional simulation software verified by historical operating data of nuclear power units. The high-pressure heater disconnection operation of nuclear power units under different seawater temperatures, turbine exhaust pressures and electrical power was simulated, thereby generating a simulation dataset for training the support vector regression model. The simulation dataset includes input features and target variables. Input features include seawater temperature, turbine exhaust pressure, and electrical power. Target variables include main steam pressure after the high-pressure heater is disconnected, main steam flow rate, extraction pressure at each stage, water-side temperature of each stage heater, and exhaust pressure. The simulation model has been validated using historical operating data of the nuclear power unit, and the relative error between the predicted and measured values ​​is controlled within ±1%. The generated simulation dataset is divided into training and testing sets according to a certain ratio; S202, Data Preprocessing The Z-score standardization method is used to standardize the input features of the support vector regression model; the formula for the Z-score standardization method is: ； in, These are standardized input feature values. These are the original input feature values. The mean of the original input feature values ​​on the training set. It is the standard deviation of the original input feature values; After training, the support vector regression model outputs standardized predicted values. These standardized predicted values ​​are then de-standardized to obtain the physical predicted values, using the following formula: ； in, These are physical predictions. It is a standardized prediction value. It is the mean of the standardized forecasts. It is the standard deviation of the standardized predicted values; S203, Training of Support Vector Regression Model The radial basis function (RBF) kernel is chosen as the default kernel function for the support vector regression model; the RBF kernel function has the following form: ； in: It is a radial basis kernel function; There are two eigenvectors and The square of the Euclidean distance between them; These are kernel function hyperparameters used to control the range of influence of the samples; S203, Hyperparameter Optimization Bayesian optimization algorithm is used to optimize the penalty hyperparameter C and kernel function hyperparameter of the support vector regression model. Automatic optimization is performed to find the optimal combination of hyperparameters (C0) that minimizes the mean squared error of the support vector regression model on a given dataset. ).

5. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 4, characterized in that, In S203, the Bayesian optimization algorithm is used to optimize the penalty hyperparameter C and kernel function hyperparameter of the support vector regression model. Automatic optimization includes the following steps: Determine the objective function: Using the mean squared error of the support vector regression model on the validation set as the objective function, the goal is to find the hyperparameter combination (C) that minimizes the mean squared error. The objective function is expressed as follows: ； in, The objective function value is represented by the mean squared error; N is the sample size. It is the predicted value of the i-th sample. It is the true value of the i-th sample; Constructing a surrogate model: A Gaussian process is used as a surrogate model to fit the hyperparameter combination (C). The probability distribution relationship between the objective function value (MSE) and the objective function value (MSE). Selection of acquisition function: The expected improvement function is adopted as the acquisition function to balance the exploration of unknown hyperparameter regions and the utilization of known hyperparameter regions that are better, and to guide the selection of the next set of hyperparameters to be evaluated. Iterative optimization: Repeat the iterative process of "selecting the next set of hyperparameters based on the acquisition function - training the support vector regression model and calculating the MSE - updating the Gaussian process surrogate model" until the preset number of iterations or convergence conditions are reached, ultimately finding the optimal combination of hyperparameters (C0) that minimizes the mean squared error of the support vector regression model on the given dataset. ).

6. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 4, characterized in that, In S3, the mechanistic model is fused with the support vector regression model, including the following steps: S301, Prediction of key operating parameters of the unit's secondary circuit thermal system after the high-pressure heater is disconnected. Real-time data on the unit’s current normal operating conditions is acquired, standardized according to the method in S202, and then input into the support vector regression model trained in S2. The support vector regression model will output standardized predicted values ​​of key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected; The standardized predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected are de-standardized by the support vector regression model to obtain the physical predicted values ​​of key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected. S302, Mechanism Model-Driven and State Calculation The physical prediction values ​​of key operating parameters of the secondary loop thermal system of the unit after the high-pressure heater is disconnected are used as input to the mechanism model established in S1. The mechanism model is driven to complete the iterative calculation of the performance of the entire secondary loop thermal system of the unit and output the performance index of the secondary loop thermal system of the unit after the high-pressure heater is disconnected.

7. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 1, characterized in that, The input to the support vector regression model is the normal operating condition data of the unit, including seawater temperature, turbine exhaust pressure and electrical power; the output of the support vector regression model is the key operating parameters of the unit's secondary loop thermal system after the high-pressure heater is disconnected, including main steam flow rate, main steam pressure, water-side inlet and outlet temperatures of each stage heater, extraction steam pressure and exhaust steam pressure of each stage heater.

8. The simulation calculation method for the disconnection condition of the high-pressure heater in a nuclear power unit according to claim 1, characterized in that, The performance indicators of the secondary loop thermal system of the unit after the high-pressure heater is disconnected include turbine power, reactor power change trend, main steam flow, parameters of each heater and condenser vacuum.

9. A computer device comprising a memory and a processor, wherein the memory stores computer-readable instructions, characterized in that, When the processor executes the computer-readable instructions, it implements the steps of the simulation calculation method for the disconnection condition of the high-pressure heater of the nuclear power unit as described in any one of claims 1-8.

10. A computer-readable storage medium storing computer-readable instructions thereon, characterized in that, When the computer-readable instructions are executed, they implement the steps of the simulation calculation method for the disconnection condition of the high-pressure heater of a nuclear power unit as described in any one of claims 1-8.

Images