An efficient nuclear reactor safety margin assessment method, system, terminal equipment and computer program product

Abstract

The present application is applicable to the field of nuclear reactor safety technology in nuclear power plants, and provides an efficient nuclear reactor safety margin assessment method, system, terminal device, and computer program product, including: constructing a training data set based on a sample data set of uncertainty parameters and time series data of key safety parameters; training an initial proxy model based on the training data set to obtain a target proxy model; efficiently calculating the sequence exceedance probability of the key safety parameters through the target proxy model; calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability and using it as the safety margin assessment result of the nuclear reactor. The method provided in the present application achieves efficient, rapid, and accurate assessment of the safety margin of the nuclear reactor by efficiently calculating the sequence exceedance probability of the key safety parameters and calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability as the safety margin assessment result of the nuclear reactor.

Description

Technical Field

[0001] The present application belongs to the technical field of safety analysis of nuclear reactors in nuclear power plants, and in particular relates to an efficient nuclear reactor safety margin assessment method, system, terminal equipment, and computer program product. Background Art

[0002] Against the backdrop of the vigorous development of "dual carbon" and clean energy, nuclear safety is the cornerstone of the sustainable development of nuclear energy. Ensuring reactor safety and reliability, identifying potential risks, and providing early warning are crucial for the flexible operation and accident response of nuclear power plants. In recent years, the Risk-Informed Safety Margin Analysis (RISMC) method, as an advanced nuclear safety assessment tool, has gradually become a key means of addressing complex operating environments and accident scenarios.

[0003] Currently, RISMC research focuses primarily on reactor design and severe accident scenarios, such as reactor power increases, nuclear power plant life extensions, and loss of coolant accidents, aiming to more realistically assess safety margins and related safety measures. However, in its application, the RISMC method still faces challenges such as high computational resource consumption, complex calculations, and long safety margin assessment cycles, making it difficult to meet the needs of nuclear power plants for rapid risk assessment and strategic decision-making during their operational phase.

[0004] Therefore, there is an urgent need for an efficient nuclear reactor safety margin assessment method to improve the margin assessment efficiency while ensuring prediction accuracy. Summary of the Invention

[0005] In view of this, embodiments of the present application provide an efficient nuclear reactor safety margin assessment method, system, terminal device, and computer program product to achieve efficient, rapid, and accurate assessment of the nuclear reactor safety margin.

[0006] A first aspect of an embodiment of the present application provides a method for evaluating a safety margin of an efficient nuclear reactor, comprising:

[0007] Construct a training dataset based on the sample dataset of uncertainty parameters and the time series data of key safety parameters;

[0008] Training an initial proxy model based on the training data set to obtain a target proxy model;

[0009] Using the target proxy model to efficiently calculate the sequence exceedance probability of the key safety parameter;

[0010] The shutdown risk of the nuclear reactor is calculated based on the sequence exceedance probability and used as a safety margin assessment result of the nuclear reactor.

[0011] In an implementation of the first aspect, efficiently calculating the sequence exceedance probability of the key safety parameter by using the target proxy model includes:

[0012] Sampling the uncertainty parameter to obtain a sample set, where the sample set includes a plurality of samples;

[0013] Calculating a key security parameter value corresponding to each sample in the sample data set by using the target proxy model;

[0014] A probability distribution of key safety parameters is formed based on the key safety parameter values, and a sequence exceedance probability of the safety parameters is calculated.

[0015] In an implementation of the first aspect, calculating the shutdown risk of a nuclear reactor based on the sequence exceedance probability and using the calculated risk as a safety margin assessment result of the nuclear reactor includes:

[0016] The reactor shutdown risk SDF is calculated based on the following formula:

[0017]

[0018] in, is the probability of abnormal event i occurring, represents the kth sequence under abnormal event i, It is a dynamic evolution sequence The probability of occurrence, Is an abnormal event i that causes any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range;

[0019] The reactor shutdown risk is used as a nuclear reactor safety margin assessment result.

[0020] In an implementation of the first aspect, after training the initial proxy model based on the training dataset to obtain the target proxy model, the method further includes:

[0021] Selecting one or more indicators of mean square error, mean absolute error, mean absolute percentage error, coefficient of determination, and error exceedance rate based on the characteristics of key safety parameters to evaluate the applicability of the target surrogate model to nuclear reactor safety margin assessment;

[0022] If the applicability of the target proxy model for nuclear reactor safety margin assessment is less than a preset applicability threshold, the target proxy model is further iteratively optimized until the applicability of the target proxy model for nuclear reactor safety margin assessment is not less than the preset applicability threshold.

[0023] In an implementation of the first aspect, constructing a training dataset based on a sample dataset of uncertainty parameters and time series data of key safety parameters further includes:

[0024] Sampling the selected uncertainty parameters to generate an uncertainty sample set, wherein the uncertainty sample set includes multiple groups of sampled samples, and each group of sampled samples includes a value of each uncertainty parameter sampled at the same time;

[0025] Based on a self-programmed simulation driver, each group of sampled samples in the uncertainty sample set is embedded in batches into an input card corresponding to a target simulation model;

[0026] running the input cards in batches to obtain large-scale simulation results;

[0027] Extracting time series extreme value information from the large-scale simulation results, the time series extreme value information including a time series extreme value of at least one key safety parameter, wherein one piece of the time series extreme value information corresponds one-to-one to a group of the sampled samples;

[0028] Each group of sample samples in the plurality of groups of sample samples is associated with the time series extreme value information of the corresponding simulation output and stored to construct a training data set.

[0029] Selecting system variables whose influence on the nuclear reactor system behavior and safety margin is greater than a preset influence threshold as uncertainty parameters;

[0030] The intervention control variable is selected as the uncertainty parameter.

[0031] In an implementation of the first aspect, the method further includes:

[0032] Select target mission scenarios based on the frequency and potential consequences of event sequences;

[0033] Establishing an initial simulation model for the target mission scenario;

[0034] simulating a target event sequence based on the initial simulation model to generate high-fidelity simulation data;

[0035] Verifying the accuracy of the initial simulation model based on the high-fidelity simulation data;

[0036] If the accuracy of the initial simulation model is lower than the preset accuracy, returning to the step of simulating the target event sequence based on the initial simulation model to generate high-fidelity simulation data;

[0037] If the accuracy of the initial simulation model is not lower than the preset accuracy, the initial simulation model is used as the target simulation model.

[0038] A second aspect of an embodiment of the present application provides an efficient nuclear reactor safety margin assessment system, comprising:

[0039] A training dataset module is used to construct a training dataset based on a sample dataset of uncertainty parameters and time series data of key safety parameters;

[0040] A proxy model training module, configured to train an initial proxy model based on the training data set to obtain a target proxy model;

[0041] A sequence exceedance probability module, configured to efficiently calculate the sequence exceedance probability of the key safety parameters using the target proxy model;

[0042] A safety margin calculation module is used to calculate the shutdown risk of the nuclear reactor based on the sequence limit violation probability and use the calculated risk as the safety margin evaluation result of the nuclear reactor.

[0043] A third aspect of an embodiment of the present application provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method for evaluating the safety margin of a high-efficiency nuclear reactor as described in the first aspect are implemented.

[0044] A fourth aspect of an embodiment of the present application provides a computer program product, including a computer program, which, when executed, enables the high-efficiency nuclear reactor safety margin assessment method as described in the first aspect to be executed.

[0045] The beneficial effect of the first aspect of the embodiment of the present application is: a training data set is constructed by constructing a sample data set based on uncertainty parameters and time series data of key safety parameters, and an initial proxy model is trained based on the training data set to obtain a target proxy model, and then the target proxy model is used to efficiently calculate the sequence exceedance probability of the key safety parameters, and finally the shutdown risk of the nuclear reactor is calculated based on the sequence exceedance probability and used as the safety margin assessment result of the nuclear reactor. By efficiently calculating the sequence exceedance probability of the key safety parameters and calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability as the safety margin assessment result of the nuclear reactor, an efficient, fast and accurate assessment of the safety margin of the nuclear reactor is achieved.

[0046] It can be understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0048] Figure 1 1 is a schematic diagram of an implementation flow of a high-efficiency nuclear reactor safety margin assessment method provided in an embodiment of the present application;

[0049] Figure 2 1 is a schematic diagram of an implementation flow of a high-efficiency nuclear reactor safety margin assessment method provided in an embodiment of the present application;

[0050] Figure 3 is a sample size variation curve diagram of the exceedance probability and its Wilson confidence interval provided in an embodiment of the present application;

[0051] Figure 4 1 is a schematic diagram of an implementation flow of a high-efficiency nuclear reactor safety margin assessment method provided in an embodiment of the present application;

[0052] Figure 5 is a curve showing the loss function change in the model training provided in the embodiment of the present application;

[0053] Figure 6 This is a prediction result diagram of the Transformer target proxy model provided in an embodiment of the present application;

[0054] Figure 7 This is a graph of the prediction results of the support vector machine target proxy model provided in an embodiment of the present application;

[0055] Figure 8 This is a prediction result diagram of the multi-layer perceptron target proxy model provided in an embodiment of the present application;

[0056] Figure 9 This is the key safety parameter distribution when the power drops to 60% and the valve opening drops to 40% at 50% water flow provided in the embodiment of the present application;

[0057] Figure 10 This is the key safety parameter distribution when the power drops to 70% and the valve opening drops to 40% at 50% water flow provided in the embodiment of the present application;

[0058] Figure 11 This is the key safety parameter distribution when the power drops to 90% and the valve opening drops to 40% at 70% water flow provided in the embodiment of the present application;

[0059] Figure 12This is the key safety parameter distribution when the power drops to 80% and the valve opening drops to 40% at 70% water flow provided in the embodiment of the present application;

[0060] Figure 13 This is the shutdown event tree at 70% feedwater flow provided by the embodiment of the present application;

[0061] Figure 14 This is the shutdown event tree at 50% feedwater flow provided by the embodiment of the present application;

[0062] Figure 15 Schematic diagram of the structure of a high-efficiency nuclear reactor safety margin assessment system provided in an embodiment of the present application;

[0063] Figure 16 is a schematic diagram of a computer device provided in an embodiment of the present application;

[0064] Figure 17 It is a schematic diagram of a computer program product provided in an embodiment of the present application. DETAILED DESCRIPTION

[0065] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0066] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0067] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0068] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0069] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0070] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0071] The present application provides an efficient nuclear reactor safety margin assessment method for achieving efficient and rapid nuclear reactor safety margin assessment. The method provided in the present application constructs a training data set by constructing a sample data set based on uncertainty parameters and time series data of key safety parameters, and trains an initial proxy model based on the training data set to obtain a target proxy model. Then, based on the target proxy model, the sequence exceedance probability of the key safety parameters is efficiently calculated, and finally, the shutdown risk of the nuclear reactor is calculated based on the sequence exceedance probability and used as the safety margin assessment result of the nuclear reactor. By efficiently calculating the sequence exceedance probability of the key safety parameters and calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability as the safety margin assessment result of the nuclear reactor, an efficient, rapid and accurate assessment of the safety margin of the nuclear reactor is achieved.

[0072] The efficient nuclear reactor safety margin assessment method provided in the embodiments of the present application can be applied to computer devices such as mobile phones, tablet computers, wearable devices, augmented reality (AR) / virtual reality (VR) devices, laptops, ultra-mobile personal computers (UMPCs), netbooks, personal digital assistants (PDAs), desktop computers, notebooks, handheld computers, and cloud servers. The embodiments of the present application do not impose any restrictions on the specific types of computer devices.

[0073] In applications, the various models in the methods provided herein can be trained using machine learning. The methods provided in the embodiments of this application can be integrated into industrial application software related to nuclear power plants to facilitate industrial information and data processing, nuclear power plant-related industrial data processing, and industrial data analysis. The methods provided herein improve the accuracy of critical safety margin predictions in complex operating conditions through data-driven proxy models, thereby ensuring flexible, safe, and reliable operation.

[0074] Expected Outlook:

[0075] 1) Provide operators with rapid, quantifiable risk-based decision support, enhancing their situational awareness. This effectively identifies the operational path after an anomaly, reduces unnecessary conservatism, and reduces reliance on expert judgment.

[0076] 2) Dynamically simulate event sequences, intelligently predict and propose diverse disposal paths and new risk insights to support the formation of optimal strategies.

[0077] like Figure 1 As shown, the embodiment of the present application provides a high-efficiency nuclear reactor safety margin assessment method, comprising:

[0078] Step S1: construct a training dataset based on a sample dataset of uncertainty parameters and time series data of key safety parameters.

[0079] Step S2: training an initial proxy model based on the training data set to obtain a target proxy model.

[0080] In applications, after constructing a training dataset, select an appropriate model architecture (e.g., Transformer-based models, LSTM, SVM) based on task requirements, computational cost, and model suitability. To improve prediction accuracy, various strategies can be employed to enhance feature extraction and model adaptability, including increasing network depth, optimizing attention mechanisms, and carefully selecting an appropriate loss function (e.g., mean squared error, mean absolute error, cross entropy) based on the specific prediction task.

[0081] Data partitioning is crucial for accurately assessing a model's generalization ability. Common strategies include simple hold-out validation, k-fold cross-validation, and leave-one-out cross-validation. Hold-out validation is computationally efficient and suitable for large datasets, such as the commonly used 7:3 training / validation split. K-fold cross-validation provides robust estimates of the model's stability across different data subsets, but at a higher computational cost. The choice of data partitioning method should balance accuracy and computational resource requirements.

[0082] Optimization algorithms play a key role in ensuring efficient and stable model training. Gradient-based optimizers, such as stochastic gradient descent (SGD), or adaptive algorithms, such as Adam, are used. SGD is simple and effective, but can converge slowly or exhibit oscillations. Adaptive optimizers generally offer faster convergence and better stability, making them more popular for complex deep learning tasks. To avoid overfitting during training, various regularization and generalization enhancement techniques, such as L2 regularization, early stopping, and cross-validation, can be employed. These methods help limit model complexity, prevent overtuning of parameters, and ensure stable model performance under various operating conditions.

[0083] Finally, the optimal combination of model parameters is determined through hyperparameter optimization (e.g., grid search, random search, Bayesian optimization).

[0084] In the application, the effectiveness and robustness of the target surrogate model are verified by using an independent test dataset to confirm its suitability for nuclear safety margin assessment.

[0085] Step S3: Utilize the target proxy model to efficiently calculate the sequence exceeding probability of the key safety parameters.

[0086] Step S4: Calculate the shutdown risk of the nuclear reactor based on the sequence exceedance probability and use it as the safety margin assessment result of the nuclear reactor.

[0087] like Figure 2 As shown, in one embodiment, the step S3, using the target proxy model to efficiently calculate the sequence exceedance probability of the key safety parameter, includes:

[0088] Step S31: sampling the uncertainty parameter to obtain a sample set, where the sample set includes a plurality of samples.

[0089] Step S32: Calculate the key security parameter value corresponding to each sample in the sample data set through the target proxy model.

[0090] Step S33: forming a probability distribution of key safety parameters based on the key safety parameter values, and calculating the sequence exceeding probability of the safety parameters.

[0091] In one embodiment, the step S4 of calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability and using the calculated risk as the safety margin assessment result of the nuclear reactor includes:

[0092] Step S41, calculating the reactor shutdown risk SDF based on the following formula:

[0093]

[0094] in, is the probability of abnormal event i occurring; represents the kth sequence under abnormal event i, It is a dynamic evolution sequence The probability of occurrence, Is an abnormal event i that causes any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range.

[0095] In the application, Also called sequence excess probability ( ), is one or more key safety parameters The joint probability of exceeding the safety threshold range.

[0096] In one embodiment, event i is a partial loss of water supply event, is the probability of water loss to a specific level, .in is the probability that the reactor power drops to a specific value, where The probability of the main steam valve dropping to a specific value, that is, the dynamic evolution sequence mainly includes valve opening adjustment and reactor power adjustment, and the sequence probability is the product of the probabilities of these two sub-events; It includes the following items: the water level of steam generators 1 and 2 drops to the shutdown signal, the water level of the pressurizer changes to the shutdown signal, and the pressurizer pressure triggers the shutdown signal. Any of the above four items can be considered as exceeding the limit.

[0097] In the application, to ensure The estimated value of has a sufficient confidence level, and each simulation is regarded as a Bernoulli test: let the total sample size be n, the number of failures (i.e., exceedances) be k, and the exceedance probability be The point estimate of is p=k / n.

[0098] Assuming that the overall distribution of the probability of exceeding the limit approximately follows a normal distribution, an interval estimate of 𝑝 is performed at a given confidence level (e.g., 95%), using the Wilson confidence interval calculation based on the inversion of the Score test. The upper bound formula is:

[0099] ;

[0100] The lower bound formula is:

[0101] ;

[0102] where z is the value of the standard normal distribution at the α / 2 quantile (e.g., z ≈ 1.96 at a 95% confidence level).

[0103] Define the interval width as , the initial sample size of the target agent model training is 1000, and each iteration training increases by 1000. In order to avoid the single interval width being too low due to accidental sampling fluctuations, this application sets: when the agent model is trained for 5 consecutive iterations When the results were all below 1%, the sample size was considered sufficient. Figure 3 The convergence trend of the model output after increasing the number of sampling times is demonstrated. When the number of samples for surrogate model training exceeds 40,000, the over-limit probability value tends to be stable, and the confidence interval width is low enough to meet the credibility requirements.

[0104] Step S42: Using the reactor shutdown risk as a nuclear reactor safety margin assessment result.

[0105] In application, this application focuses on the impact of operations before the nuclear reactor shutdown on the consequences in non-accident conditions. Therefore, the calculation of the nuclear reactor safety margin is transformed into the aforementioned shutdown risk SDF calculation.

[0106] In the application, the safety margin calculation method is combined with a statistical analysis of the limit values ​​of key safety parameters. After assessing the probability of exceeding the limit, a safety parameter distribution diagram is further drawn to intuitively demonstrate the changes in safety margins under different conditions. Ultimately, the calculation results are presented through visualization methods, including the dynamic event sequence of key safety parameters, risk limit surfaces, and probability distribution curves of key safety parameters. This allows operators to intuitively understand the system status and provide data support for optimized decision-making.

[0107] In one embodiment, after the step S3, in which the initial proxy model is trained based on the training data set to obtain the target proxy model, further comprises:

[0108] Step S401 , based on the characteristics of key safety parameters, one or more indicators selected from mean square error, mean absolute error, mean absolute percentage error, coefficient of determination, and error exceedance rate are selected to evaluate the applicability of the target proxy model to nuclear reactor safety margin assessment.

[0109] In application, after obtaining the target proxy model, it is necessary to use a variety of indicators to comprehensively evaluate the prediction performance of the target proxy model to ensure that it can meet the application requirements of the safety margin analysis of nuclear power plants. In view of the specific characteristics of parameter prediction in nuclear power plant safety analysis, the selection of evaluation indicators needs to be determined in combination with the specific prediction task type. When the target proxy model is applied to the absolute numerical prediction of key safety parameters, the mean squared error (MSE) and mean absolute error (MAE) should be used to evaluate the error size to intuitively reflect the average level of prediction error. If the MAE is less than 1% of the magnitude of the original data, it means that the target proxy model has better performance. When comparing the prediction effects of parameters of different dimensions, the mean absolute percentage error (MAPE) is generally used to eliminate the influence of dimensional differences. When there are a large number of values ​​close to 0 in the data, the calculated result of MAPE may be too high. In this case, the coefficient of determination ( ) reflects the fitting effect of the target proxy model on the overall change trend of the parameters, The closer it is to 1, the stronger the model's ability to explain the trend of parameter changes. It is usually required >0.9 to ensure the reliability of the model.

[0110] The calculation formula for the above indicators is as follows:

[0111]

[0112]

[0113]

[0114]

[0115] Where n is the sample size, is the actual value, is the predicted value, is the average of the actual values.

[0116] Step S402: If the applicability of the target proxy model for nuclear reactor safety margin assessment is less than a preset applicability threshold, further iteratively optimize the target proxy model until the applicability of the target proxy model for nuclear reactor safety margin assessment is not less than the preset applicability threshold.

[0117] In applications, the generalization capability of the target surrogate model is further verified by combining it with independent test data. If the target surrogate model's performance does not meet the predetermined generalization standard, additional training samples can be added, the model structure can be optimized, or hyperparameters can be adjusted. The target surrogate model is then further iteratively optimized until its prediction accuracy and generalization performance meet the requirements of actual application scenarios. These steps enable the target surrogate model to achieve the necessary prediction accuracy and reliability required to support risk-guided decision-making and safety margin assessment in nuclear power plants.

[0118] like Figure 4 As shown, in one embodiment, step S1, constructing a training data set based on a sample data set of uncertainty parameters and time series data of key safety parameters, includes:

[0119] Step S11 , sampling the selected uncertainty parameters to generate an uncertainty sample set, wherein the uncertainty sample set includes multiple groups of sampling samples, and each group of the sampling samples includes a value of each of the uncertainty parameters sampled at the same time.

[0120] In applications, advanced sampling methods are used to generate data that reflects variations in different parameters. For example, Latin Hypercube Sampling (LHS) provides more uniform coverage of the input space than simple random sampling or Monte Carlo sampling, thereby improving training efficiency and enhancing the robustness of the resulting model. Furthermore, after sampling, all variables are normalized to a consistent scale.

[0121] Step S12: Based on the self-programmed simulation driver, each group of sampled samples in the uncertainty sample set is embedded in batches into the input card corresponding to the target simulation model.

[0122] In the application, the target simulation model is the optimized nuclear reactor thermal hydraulic simulation program.

[0123] Step S13: running the input cards in batches to obtain large-scale simulation results.

[0124] In the application, the target simulation model is used to simulate the sample data set to obtain the time series data of key safety parameters, with particular attention paid to the extreme values ​​in the time series data, such as the peak temperature of the fuel cladding, the low water level of the steam generator, and the high pressure of the regulator.

[0125] Step S14: extracting time series extreme value information from the large-scale simulation results, wherein the time series extreme value information includes a time series extreme value of at least one key safety parameter, and one piece of the time series extreme value information corresponds one-to-one to a group of the sampling samples.

[0126] Step S15 , storing each group of sample samples in the plurality of groups of sample samples in association with the time series extreme value information of the corresponding simulation output, so as to construct a training data set.

[0127] In the application, noise reduction technology is applied to improve the data quality of the training dataset and the stability of model training. Therefore, by normalizing all data to a comparable scale, the training efficiency of subsequent proxy models is improved.

[0128] In one embodiment, before performing sampling based on the uncertainty parameter to obtain a sample data set of the uncertainty parameter in step S11, the following steps may be further included:

[0129] Step S01 : selecting a system variable whose influence on the nuclear reactor system behavior and safety margin is greater than a preset influence threshold as an uncertainty parameter.

[0130] In the application, sensitivity analysis is used to identify system variables that have a significant impact on the nuclear reactor system behavior and safety margin as uncertainty parameters.

[0131] Step S02: Select an intervention control variable as an uncertainty parameter.

[0132] In application, intervention control variables include operator actions and control variables reflecting intervention options that could significantly change the course of the accident (e.g., timing of feedwater restoration, adjustment of reactor power).

[0133] After selecting the uncertainty parameters, the uncertainty of the uncertainty parameters is expressed using a probability distribution based on engineering constraints, reliability databases, or technical reports. It can be understood that the above uncertainty parameter selection process ensures that the dataset contains both the influencing system variables and the actual operator actions.

[0134] In one embodiment, the method further comprises:

[0135] Step S51 : selecting a target task scenario based on the occurrence frequency and potential consequences of the event sequence.

[0136] In the application, target mission scenarios are selected based on the frequency of occurrence and potential consequences of event sequences, and specific scenarios that require decision support during the operation of nuclear power plants or nuclear facilities are identified, with particular focus on risk management under complex conditions, with an emphasis on the following two types of scenarios: risk-important scenarios and operation-important scenarios.

[0137] Among them, risk-critical scenarios refer to events or event sequences with high risk contributions. They are ranked by risk importance based on the calculation results of importance indicators, such as Fussel-Vesely importance, risk increase importance (RAW), and risk reduction importance (RRW). Operationally critical scenarios refer to scenarios that require the main control room operator to correctly judge and execute actions and have significant consequences. Timely and effective responses by operators will not directly lead to unplanned shutdowns. Therefore, it is necessary to carry out risk-guided comprehensive decision support to provide risk warnings and action time windows before operator intervention.

[0138] Step S52: establishing an initial simulation model for the target task scenario.

[0139] In the application, the target task scenario is modeled by a thermal-hydraulic (TH) model, for example, by using a TH simulation program (such as RELAP5) to model the target task scenario to obtain an initial simulation model.

[0140] Step S53: simulating the target event sequence based on the initial simulation model to generate high-fidelity simulation data.

[0141] The simulation environment of the initial simulation model was carefully designed to accurately reflect the response of nuclear power plant (NPP) systems and equipment. This involved high-precision simulation of key system parameters such as reactor power level, coolant flow, system pressure, and temperature. Furthermore, for each selected operating condition, appropriate boundary conditions were set to balance computational accuracy and model granularity while avoiding excessive computational costs.

[0142] Step S54, verifying the accuracy of the initial simulation model based on the high-fidelity simulation data;

[0143] In application, the TH simulation program simulates a specific sequence of events for a target mission scenario, including a reduction in feedwater flow, adjustments to the valve opening of the steam generator main steam pipe, and changes in reactor power level, to observe how operator interventions affect water levels and system stability. During the simulation, key safety parameters are recorded to capture the system's transient behavior. The TH simulation results are then compared with experimental data, historical records, calculated TH simulation results from the plant safety analysis report, or theoretical expectations to verify the model's accuracy.

[0144] Step S55: If the accuracy of the initial simulation model is lower than the preset accuracy, return to step S53 to simulate the target event sequence based on the initial simulation model to generate high-fidelity simulation data.

[0145] In application, if the accuracy of the initial simulation model is lower than the preset accuracy, the process returns to step S53 to achieve continuous optimization of the initial simulation model.

[0146] Step S56: If the accuracy of the initial simulation model is not lower than the preset accuracy, the initial simulation model is used as the target simulation model.

[0147] In the application, if the accuracy of the initial simulation model is not less than the preset accuracy, the initial simulation model will be used as the target simulation model for subsequent use.

[0148] In the application, the key safety parameter distribution, the risk of each intervention path and the limit surface / risk profile can be obtained based on the target proxy model. The key safety parameter distribution is the probability distribution of the key safety parameters calculated by the target proxy model, which is used to evaluate the The risk of each intervention path is quantified by the target surrogate model as the likelihood of adverse outcomes resulting from different operational strategies, enabling operators to weigh the risk level of each path against operational objectives. The target surrogate model constructs a limit surface or risk profile that illustrates how certain parameter limits (e.g., maximum pressurizer pressure, minimum water level) interact with each intervention scenario. This highlights the boundaries between safe and unsafe areas, enabling more direct, risk-based decision-making.

[0149] This embodiment of the application also selects a rupture of a feedwater branch pipe in the secondary circuit system of a nuclear power plant (i.e., a partial loss of feedwater accident) as an example of a target task scenario, and uses the above-mentioned method provided in this application to perform a safety margin assessment. The specific process of the example is as follows.

[0150] According to nuclear power plant operating procedures, some dehydration events are not severe enough to directly trigger a reactor shutdown, but they could affect sustained power operation, posing a potential risk. Therefore, this paper simulates the dynamic evolution of an accident through thermal-hydraulic (TH) transient analysis and evaluates different operator intervention strategies to optimize operational responses, avoid unplanned reactor shutdowns, and enhance operational safety and flexibility. Using the RELAP5 nuclear reactor system thermal-hydraulic simulation program, this study constructed a thermal-hydraulic simulation model of a typical 900MW dual-circuit pressurized water reactor, including key equipment such as the reactor pressure vessel, core, hot and cold pipe sections, main pumps, steam generators, and pressurizers. The main design parameters are shown in Table 1.

[0151] Table 1 Design values ​​of main parameters of nuclear power plants

[0152]

[0153] Case scenario description

[0154] In this example, a partial loss of secondary feedwater scenario involves a leak in a steam generator's feedwater branch pipe, resulting in a reduction in feedwater flow and, in turn, a continuous drop in the steam generator's water level. To address this situation, operators can take measures such as reducing core power and adjusting the main steam pipe valve opening to maintain the steam generator water level and system stability. During the simulation, the RELAP5 program was used to analyze the evolution of key safety parameters such as steam generator water level, pressurizer pressure, and reactor power as feedwater flow changes, further exploring the effectiveness of different intervention strategies.

[0155] The main events and time nodes are as follows:

[0156] At t=1000s, the water feed branch pipe of steam generator 1 leaks, causing partial water loss in the secondary circuit system and a drop in the water level of the steam generator, affecting steam production and heat exchange capacity.

[0157] t=1060s: The operator begins to take countermeasures, including reducing the reactor power and adjusting the valve opening of the steam generator main steam pipe.

[0158] To mitigate the impact of accidents. When the pressure of the regulator exceeds the set limit, the safety valve and the relief valve will start normally to effectively reduce the system pressure and prevent overpressure failure.

[0159] To simplify the calculation, the following assumptions are made:

[0160] 1. Use the average value of the temperature feedback coefficients of the reactor fuel elements and moderator.

[0161] 2. Maintain constant vacuum in the condenser to ensure normal operation of the turbine.

[0162] 3. The reactor maintains partial power operation and the automatic shutdown is not triggered.

[0163] 4. The main steam control valve, steam discharge valve and pressure regulator safety valve are operating normally.

[0164] Key safety parameters exceeding safety thresholds may trigger an automatic reactor shutdown. According to the accident analysis report for loss of feedwater, there are three types of parameter signals that may lead to a reactor shutdown:

[0165] 1. Low water level trip signal (<1.24m) of steam generator water level (SG1lev, SG2lev).

[0166] 2. Overpressure trip signal of pressurizer pressure (PRZpre) (>16.3Mpa).

[0167] 3. Low water level trip signal of pressurizer water level (PRZlev) (<0.5m).

[0168] Target surrogate model training and validation

[0169] After the thermal hydraulic model is built, the data required to train the target agent model is generated. This study focuses on analyzing the changes in water flow and the intelligent intervention strategies under different accident sequences, mainly considering the impact of the following uncertainty parameters on accident consequences:

[0170] 1. The secondary circuit water supply loses flow level.

[0171] 2. The operator manually adjusts the reactor power value.

[0172] 3. The steam valve opening manually controlled by the operator.

[0173] 4. Recovery time of water supply flow.

[0174] To construct the dataset required for training the target surrogate model, this study first modeled a partial loss-of-water (LOFW) accident in the secondary circuit system of a nuclear power plant and defined key operational events and corresponding operator interventions. The operator's intervention strategy primarily consisted of adjusting the reactor power level and the main steam outlet valve opening. To fully cover possible intervention paths, these operational variables were discretized, generating a total of 32 different intervention paths. Furthermore, varying feedwater flow rates and their recovery times also influenced the accident evolution. Therefore, within each intervention path, Latin hypercube sampling (LHS) was used to randomly sample these two parameters to expand the input variable space and enhance the model's generalization ability. Ultimately, 50 samples were sampled for each intervention path, and 50 samples were sampled for the accident path where no intervention was taken, generating a total of 1,650 data samples. Table 2 details the sampling method for operational events and interventions.

[0175] Table 2 Operational events and intervention measures

[0176]

[0177] After data sampling, all samples were calculated using the RELAP5 thermal-hydraulic simulation program to obtain time series data for key safety parameters (such as steam generator water level and pressurizer pressure). To train the target surrogate model, the maximum values ​​of these key safety parameters were extracted to determine whether an over-limit condition had occurred and to further calculate the safety margin. During the target surrogate model training process, feedwater flow, feedwater recovery time, and operator intervention measures were selected as input parameters. Key safety parameters (i.e., Steam Generator 1 Level (SG1 Lev), Steam Generator 2 Level (SG2 Lev), Pressurizer Pressure (PRZ Pre), and Pressurizer Level (PRZ Lev)) were selected as output parameters to ensure the model accurately predicts safety margin levels under different operating conditions. Table 3 lists the input and output parameters of the target surrogate model.

[0178] Table 3 Input parameters and output parameters of target proxy model

[0179]

[0180] The collected data was then divided proportionally into a training set (70%), a test set (15%), and a validation set (15%) to ensure the model's strong learning and generalization capabilities, as well as stable performance evaluation. During training, the mean squared error (MSE) was used as the loss function, and the Adam optimizer was used to update model weights. A dropout mechanism was also introduced to reduce the risk of overfitting. Figure 5 The following figure shows the trend of the model training loss function. After 60 epochs, the loss gradually converged and fell below 1%, indicating that the model has achieved excellent fit. Because this study involved relatively few input scenarios and parameter dimensions, the model training efficiency was high, completing within 150 epochs with low overall computing resource consumption (approximately 1 minute), meeting the requirements for efficient computing.

[0181] After training is completed, the performance of the target proxy model based on Transformer is verified. MSE, MAE, MAPE and As an evaluation indicator. Considering that the parameters involved in the case are 4-dimensional and the parameter interval span is large, directly calculating MSE and MAE will lead to insignificant errors in some dimensions, so the calculated data in this case is normalized. In addition, since there are a large number of 0 values ​​in the steam generator water level data, this will cause the result to be too large when calculating MAPE, and it cannot actually reflect the accuracy of the model. Finally, this case sets the determination coefficient >0.95 is used as a criterion to ensure that the target surrogate model has high prediction accuracy and generalization ability and can meet the needs of actual engineering applications. Figures 6 to 8 The following are the regression performances of the Transformer model, Support Vector Machine (SVM), and Multi-Layer Perceptron (MLP), respectively. As can be seen from the figure, the Transformer model performs better than the traditional regression models (SVM and MLP) in regression tasks. This is because the Transformer can effectively capture the global patterns in complex data through the self-attention mechanism, making the predicted value closer to the actual value. Table 4 further verifies the advantage of the Transformer in prediction accuracy. The comparison results show that compared with MLP and SVM, the MSE after combining the Transformer algorithm decreased by about 89.5% and 90.5%, and the MAE decreased by about 71.5% and 69.2%. It was increased to 0.982, and the accuracy and stability of regression calculations were greatly improved.

[0182] Table 4 Comparison of performance indicators of different regression models

[0183]

[0184] Water restoration times for different intervention paths were sampled, and a regression model (target surrogate model) calculated the corresponding key output parameter values ​​and statistically analyzed the probability of exceeding thresholds. Based on a comprehensive analysis of these paths, several typical critical paths were extracted at two water flow levels. A shutdown event tree was then constructed to comprehensively evaluate the system's safety margin under various scenarios. Figures 9 to 12 The kernel density estimation (KDE) distribution function of the key safety parameters of the shutdown under different intervention measures was further revealed. Figure 9 The key safety parameter distribution when the power drops to 60% and the valve opening drops to 40% at 50% feedwater flow rate; Figure 10 The key safety parameter distribution when the power drops to 70% and the valve opening drops to 40% at 50% water flow rate; Figure 11 The key safety parameter distribution when the power drops to 90% and the valve opening drops to 40% at 70% water flow rate; Figure 12 The distribution of key safety parameters when the power drops to 80% and the valve opening drops to 40% at a feedwater flow rate of 70%. Clearly, at a feedwater flow rate of 70%, rationally adjusting the reactor power and precisely controlling the valve opening can effectively avoid unplanned shutdowns. However, at a feedwater flow rate of 50%, even with appropriate intervention measures, shutdown signals can still be triggered in some scenarios. Figure 13 and Figure 14 The shutdown event tree shows the probability of sequence exceeding the limit under different operating conditions. Figure 13 This is the shutdown event tree at 70% feedwater flow rate; Figure 14 This is the shutdown event tree for a 50% feedwater flow rate, where success indicates a successful avoidance of a reactor shutdown, and failure indicates an emergency shutdown. Combined with the probabilistic safety analysis data of the specific power plant (the probability of a loss of feedwater event and the probability of a head event), the shutdown risk is further calculated using the reactor shutdown risk calculation formula. In this example, event i is a partial loss of feedwater event. is the probability of loss of feedwater to a specific level; represents the kth sequence under the partial loss of water supply event, It is a dynamic evolution sequence The probability that in a specific case .in is the probability that the reactor power drops to a specific value, where The probability of the main steam valve dropping to a specific value, that is, the dynamic evolution sequence mainly includes valve opening adjustment and reactor power adjustment, and the sequence probability is the product of the probabilities of these two sub-events; Is an abnormal event i that causes any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range, in the example This includes the following: the water level in steam generators 1 and 2 drops to a shutdown signal, the pressurizer water level changes to a shutdown signal, and the pressurizer pressure triggers a shutdown signal. Any of these four conditions is considered an over-limit condition. This validates the applicability of the proposed model and method in multi-scenario assessments.

[0185] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0186] The present application also provides a high-efficiency nuclear reactor safety margin assessment system for performing the steps of the high-efficiency nuclear reactor safety margin assessment method described above. The high-efficiency nuclear reactor safety margin assessment system can be a virtual appliance in a computer device, executed by a processor of the computer device, or can be the computer device itself.

[0187] like Figure 15 As shown, the efficient nuclear reactor safety margin assessment system 50 provided in the embodiment of the present application includes:

[0188] A training data set module 501 is used to construct a training data set based on a sample data set of uncertainty parameters and time series data of key safety parameters;

[0189] A proxy model training module 502 is configured to train an initial proxy model based on the training data set to obtain a target proxy model;

[0190] A sequence exceeding probability module 503 is used to efficiently calculate the sequence exceeding probability of the key safety parameter using the target proxy model;

[0191] The safety margin calculation module 504 is configured to calculate the shutdown risk of the nuclear reactor based on the sequence exceedance probability and use the calculated risk as the safety margin assessment result of the nuclear reactor.

[0192] In one embodiment, the sequence exceedance probability module 503 is configured to:

[0193] Sampling the uncertainty parameter to obtain a sample set, where the sample set includes a plurality of samples;

[0194] Calculating a key security parameter value corresponding to each sample in the sample data set by using the target proxy model;

[0195] A probability distribution of key safety parameters is formed based on the key safety parameter values, and a sequence exceedance probability of the safety parameters is calculated.

[0196] In one embodiment, the safety margin calculation module 504 is configured to:

[0197] The reactor shutdown risk SDF is calculated based on the following formula:

[0198]

[0199] in, is the probability of abnormal event i occurring It is a dynamic evolution sequence The probability of occurrence is the abnormal event i causing any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range.

[0200] The reactor shutdown risk is used as a nuclear reactor safety margin assessment result.

[0201] In one embodiment, the high-efficiency nuclear reactor safety margin assessment system 50 further includes:

[0202] Proxy model validation module for:

[0203] Selecting one or more indicators of mean square error, mean absolute error, mean absolute percentage error, coefficient of determination, and error exceedance rate based on the characteristics of key safety parameters to evaluate the applicability of the target surrogate model to nuclear reactor safety margin assessment;

[0204] If the applicability of the target proxy model for nuclear reactor safety margin assessment is less than a preset applicability threshold, the target proxy model is further iteratively optimized until the applicability of the target proxy model for nuclear reactor safety margin assessment is not less than the preset applicability threshold.

[0205] In one embodiment, the training dataset module 501 is used to:

[0206] Sampling the selected uncertainty parameters to generate an uncertainty sample set, wherein the uncertainty sample set includes multiple groups of sampled samples, and each group of sampled samples includes a value of each uncertainty parameter sampled at the same time;

[0207] Based on a self-programmed simulation driver, each group of sampled samples in the uncertainty sample set is embedded in batches into an input card corresponding to a target simulation model;

[0208] running the input cards in batches to obtain large-scale simulation results;

[0209] Extracting time series extreme value information from the large-scale simulation results, the time series extreme value information including a time series extreme value of at least one key safety parameter, wherein one piece of the time series extreme value information corresponds one-to-one to a group of the sampled samples;

[0210] Each group of sample samples in the plurality of groups of sample samples is associated with the time series extreme value information of the corresponding simulation output and stored to construct a training data set.

[0211] In one embodiment, the high-efficiency nuclear reactor safety margin assessment system 50 further includes:

[0212] Uncertainty parameter screening module, used for:

[0213] Selecting system variables whose influence on the nuclear reactor system behavior and safety margin is greater than a preset influence threshold as uncertainty parameters;

[0214] The intervention control variable is selected as the uncertainty parameter.

[0215] In one embodiment, the high-efficiency nuclear reactor safety margin assessment system 50 further includes:

[0216] Target simulation model module for:

[0217] Select target mission scenarios based on the frequency and potential consequences of event sequences;

[0218] Establishing an initial simulation model for the target mission scenario;

[0219] simulating a target event sequence based on the initial simulation model to generate high-fidelity simulation data;

[0220] Verifying the accuracy of the initial simulation model based on the high-fidelity simulation data;

[0221] If the accuracy of the initial simulation model is lower than the preset accuracy, returning to the step of simulating the target event sequence based on the initial simulation model to generate high-fidelity simulation data;

[0222] If the accuracy of the initial simulation model is not lower than the preset accuracy, the initial simulation model is used as the target simulation model.

[0223] In application, each module in the efficient nuclear reactor safety margin assessment system may be a software program module, or may be implemented through different logic circuits integrated in a processor, or may be implemented through multiple distributed processors.

[0224] Figure 16 This is a schematic diagram of the structure of a computer device provided in one embodiment of the present application. Figure 16 As shown, the computer device 6 of this embodiment includes: at least one processor 60 ( Figure 16 Only one is shown in the figure) a processor, a memory 61, and a computer program 62 stored in the memory 61 and executable on the at least one processor 60, wherein the processor 60 implements the steps of any of the above-mentioned embodiments of the method for evaluating the safety margin of a high-efficiency nuclear reactor when executing the computer program 62.

[0225] The computer device may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art will appreciate that Figure 16This is merely an example of the computer device 6 and does not constitute a limitation on the computer device 6 . The computer device 6 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device 6 may also include input and output devices, network access devices, etc.

[0226] The processor 60 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor.

[0227] In some embodiments, the memory 61 may be an internal storage unit of the computer device 6, such as a hard drive or memory of the computer device 6. In other embodiments, the memory 61 may also be an external storage device of the computer device 6, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the computer device 6. Furthermore, the memory 61 may include both an internal storage unit of the computer device 6 and an external storage device. The memory 61 is used to store an operating system, application programs, a boot loader, data, and other programs, such as the program code of the computer program. The memory 61 may also be used to temporarily store data that has been output or is about to be output.

[0228] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of this application. Their specific functions and technical effects can be found in the method embodiment section and will not be repeated here.

[0229] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0230] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned various method embodiments can be implemented.

[0231] like Figure 17 As shown, an embodiment of the present application provides a computer program product 7, including a computer program 62. When the computer program 62 is run, the steps in the above-mentioned embodiments of the high-efficiency nuclear reactor safety margin assessment method are executed.

[0232] If the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process steps in the above-mentioned method embodiments by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a device / terminal equipment, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, removable hard drives, magnetic disks, or optical disks. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electric carrier signals or telecommunication signals.

[0233] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0234] In the embodiments provided in this application, it should be understood that the disclosed computer devices and methods can be implemented in other ways. For example, the computer device embodiments described above are merely illustrative. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0235] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A method for evaluating safety margin of a high-efficiency nuclear reactor, characterized in that: include: Construct a training dataset based on the sample dataset of uncertainty parameters and the time series data of key safety parameters; Training an initial proxy model based on the training data set to obtain a target proxy model; Using the target proxy model to efficiently calculate the sequence exceedance probability of the key safety parameter; Calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability and using the calculated risk as a safety margin assessment result of the nuclear reactor; The method of efficiently calculating the sequence exceeding limit probability of the key safety parameter by using the target proxy model includes: Sampling the uncertainty parameter to obtain a sample set, where the sample set includes a plurality of samples; Calculating a key security parameter value corresponding to each sample in the sample data set by using the target proxy model; forming a probability distribution of key safety parameters based on the key safety parameter values, and calculating a sequence exceedance probability of the safety parameters; The calculating the shutdown risk of the nuclear reactor based on the sequence exceedance probability and using the calculated risk as the safety margin assessment result of the nuclear reactor includes: The reactor shutdown risk SDF is calculated based on the following formula: in, is the probability of abnormal event i occurring, represents the kth sequence under abnormal event i, It is a dynamic evolution sequence The probability of occurrence, Is an abnormal event i that causes any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range; The reactor shutdown risk is used as a nuclear reactor safety margin assessment result.

2. The method for evaluating the safety margin of a high-efficiency nuclear reactor according to claim 1, wherein: After the initial proxy model is trained based on the training data set to obtain the target proxy model, the method further includes: Selecting one or more indicators of mean square error, mean absolute error, mean absolute percentage error, coefficient of determination, and error exceedance rate based on the characteristics of key safety parameters to evaluate the applicability of the target surrogate model to nuclear reactor safety margin assessment; If the applicability of the target proxy model for nuclear reactor safety margin assessment is less than a preset applicability threshold, the target proxy model is further iteratively optimized until the applicability of the target proxy model for nuclear reactor safety margin assessment is not less than the preset applicability threshold.

3. The method for evaluating the safety margin of a high-efficiency nuclear reactor according to claim 1, wherein: The training data set is constructed based on the sample data set of uncertainty parameters and the time series data of key safety parameters, including: Sampling the selected uncertainty parameters to generate an uncertainty sample set, wherein the uncertainty sample set includes multiple groups of sampled samples, and each group of sampled samples includes a value of each uncertainty parameter sampled at the same time; Based on a self-programmed simulation driver, each group of sampled samples in the uncertainty sample set is embedded in batches into an input card corresponding to a target simulation model; running the input cards in batches to obtain large-scale simulation results; Extracting time series extreme value information from the large-scale simulation results, the time series extreme value information including a time series extreme value of at least one key safety parameter, wherein one piece of the time series extreme value information corresponds one-to-one to a group of the sampled samples; Each group of sample samples in the plurality of groups of sample samples is associated with the time series extreme value information of the corresponding simulation output and stored to construct a training data set.

4. The method for evaluating the safety margin of a high-efficiency nuclear reactor according to claim 3, wherein: Before sampling based on the uncertainty parameter to obtain a sample data set of the uncertainty parameter, the method further includes: Selecting system variables whose influence on the nuclear reactor system behavior and safety margin is greater than a preset influence threshold as uncertainty parameters; The intervention control variable is selected as the uncertainty parameter.

5. The method for evaluating the safety margin of a high-efficiency nuclear reactor according to claim 3, wherein: The method further comprises: Select target mission scenarios based on the frequency and potential consequences of event sequences; Establishing an initial simulation model for the target mission scenario; simulating a target event sequence based on the initial simulation model to generate high-fidelity simulation data; Verifying the accuracy of the initial simulation model based on the high-fidelity simulation data; If the accuracy of the initial simulation model is lower than the preset accuracy, returning to the step of simulating the target event sequence based on the initial simulation model to generate high-fidelity simulation data; If the accuracy of the initial simulation model is not lower than the preset accuracy, the initial simulation model is used as the target simulation model.

6. An efficient nuclear reactor safety margin assessment system, characterized in that: include: A training dataset module is used to construct a training dataset based on a sample dataset of uncertainty parameters and time series data of key safety parameters; A proxy model training module, configured to train an initial proxy model based on the training data set to obtain a target proxy model; A sequence exceedance probability module, configured to efficiently calculate the sequence exceedance probability of the key safety parameters using the target proxy model; The method of efficiently calculating the serial exceedance probability of the key safety parameter using the target proxy model includes: sampling the uncertainty parameter to obtain a sample set, wherein the sample set includes a plurality of samples; calculating the key safety parameter value corresponding to each sample in the sample data set using the target proxy model; forming a probability distribution of the key safety parameter based on the key safety parameter values, and calculating the serial exceedance probability of the safety parameter; A safety margin calculation module is configured to calculate a shutdown risk of a nuclear reactor based on the sequence exceedance probability and use the calculated risk as a safety margin assessment result of the nuclear reactor; the calculation of the shutdown risk of the nuclear reactor based on the sequence exceedance probability and using the calculated risk as a safety margin assessment result of the nuclear reactor includes: calculating a reactor shutdown risk SDF based on the following formula: in, is the probability of abnormal event i occurring, represents the kth sequence under abnormal event i, It is a dynamic evolution sequence The probability of occurrence, Is an abnormal event i that causes any one or more key safety parameters The probability of exceeding the safety threshold range, Key safety parameters The probability of being above the safety threshold range; and taking the reactor shutdown risk as the nuclear reactor safety margin assessment result.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the high-efficiency nuclear reactor safety margin assessment method according to any one of claims 1 to 5 are implemented.

8. A computer program product, characterized in that The invention comprises a computer program, which enables the efficient nuclear reactor safety margin assessment method according to any one of claims 1 to 5 to be executed when the computer program is executed.

Images