Nuclear power station operation optimization method and system, equipment medium and product

Abstract

The invention provides a nuclear power station operation optimization method and system, an equipment medium and a product, and relates to the field of nuclear power, the optimization method comprises the steps that actual safety analysis data of a nuclear power station and simulation safety analysis data of a nuclear power simulation model under the same operation data are acquired, and the nuclear power simulation model is obtained based on all-condition simulation of the nuclear power station; determining state data of the nuclear power station based on the actual safety analysis data and the simulated safety analysis data; and in combination with the actual safety analysis data, the simulation safety analysis data and the state data, generating a corresponding optimization strategy which is used for optimizing the operation process of the nuclear power station. Through the actual safety analysis data and the simulation analysis data, the operation optimization strategy of the nuclear power station can be generated, accurate working condition simulation and optimization decision can be performed in the whole life cycle of the nuclear power station, and the emergency response capability of workers is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of nuclear power, and in particular to a method, system, equipment medium and product for optimizing the operation of a nuclear power plant. Background Art

[0002] During the operation of a nuclear power plant, adaptive full-operating-condition simulation of the nuclear power plant is an important tool for nuclear power plant operation management. It can truly restore the dynamic operating conditions of the nuclear power plant, including normal, abnormal and accident conditions, and provide support for staff training, design verification and operation optimization.

[0003] However, current traditional nuclear power plant simulations are primarily used for static training and offline analysis, and are unable to manage nuclear power plant operations. This makes it difficult to meet the needs of real-time optimization during nuclear power plant operations. Summary of the Invention

[0004] The technical problem to be solved by the present disclosure is to overcome the defect in the prior art that traditional nuclear power plant simulation is difficult to meet the demand for real-time optimization during the operation of the nuclear power plant, and to provide an optimization method, system, equipment medium and product for the operation of a nuclear power plant.

[0005] The present disclosure solves the above technical problems through the following technical solutions:

[0006] In a first aspect, an embodiment of the present disclosure provides a method for optimizing the operation of a nuclear power plant, the method comprising:

[0007] Obtaining actual safety analysis data of the nuclear power plant and simulated safety analysis data of the nuclear power simulation model under the same operating data. The nuclear power simulation model is based on a full-operation simulation of the nuclear power plant;

[0008] Determine status data of the nuclear power plant based on actual safety analysis data and simulated safety analysis data;

[0009] By combining actual safety analysis data, simulated safety analysis data and status data, corresponding optimization strategies are generated, which are used to optimize the operation process of the nuclear power plant.

[0010] Optionally, the optimization method further includes a training step of a nuclear power simulation model;

[0011] The training steps include:

[0012] Inputting operational data into pre-built nuclear power simulation models;

[0013] Construct an objective function based on the output of the nuclear power simulation model and actual safety analysis data;

[0014] With the goal of minimizing the objective function, the parameters of the nuclear power simulation model are adjusted through a hybrid optimization algorithm to obtain a trained nuclear power simulation model. The hybrid optimization algorithm is constructed based on the particle swarm optimization algorithm and the neural network algorithm.

[0015] Optionally, with the objective of minimizing the objective function, the parameters of the nuclear power simulation model are adjusted by a hybrid optimization algorithm to obtain a trained nuclear power simulation model, including:

[0016] Generate an initial population, which includes multiple particles, and one particle corresponds to a candidate solution set of the parameters;

[0017] In each round of iteration, the fitness of each particle in the population is calculated, the best particle in the population is updated, and the position and velocity of each particle in the population are updated until the sum of the fitness of all particles in the population is less than the preset threshold;

[0018] In each round of iteration, a loss function is constructed based on the output of the nuclear power simulation model and actual safety analysis data. The output of the nuclear power simulation model is adjusted through a neural network algorithm with the goal of minimizing the gradient of the loss function.

[0019] The optimal position of each particle in the population is taken as the global optimal solution of the parameters to obtain a trained nuclear power simulation model.

[0020] Optionally, the status data includes change data of the safety analysis data of the nuclear power plant within a preset time period;

[0021] Determine the status data of the nuclear power plant based on actual safety analysis data and simulated safety analysis data, including:

[0022] Determine the change data of the safety analysis data of the nuclear power plant within a preset time period based on the actual safety analysis data and the simulated safety analysis data;

[0023] Combine actual safety analysis data, simulated safety analysis data, and status data to generate corresponding optimization strategies, including:

[0024] Combine actual security analysis data, simulated security analysis data and change data to generate corresponding optimization strategies.

[0025] Optionally, the status data also includes fault data;

[0026] After determining the change data of the safety analysis data of the nuclear power plant within a preset time period based on the actual safety analysis data and the simulated safety analysis data, the following steps are included:

[0027] Determine fault data based on actual safety analysis data, simulated safety analysis data, and change data;

[0028] Generate corresponding optimization strategies by combining actual safety analysis data, simulated safety analysis data, and status data, including:

[0029] Generate corresponding optimization strategies by combining actual safety analysis data, simulated safety analysis data, and fault data.

[0030] Optionally, the fault data includes fault content and fault cause;

[0031] Determine fault data based on actual safety analysis data, simulated safety analysis data, and change data, including:

[0032] Determine the fault content based on actual safety analysis data, simulated safety analysis data, and change data;

[0033] Take the fault content as the target event, and at least one of the actual safety analysis data, simulated safety analysis data, and change data as the status, and analyze through a dynamic fault tree to determine the fault cause.

[0034] In a second aspect, an embodiment of the present disclosure provides an optimization system for nuclear power plant operation, and the optimization system includes:

[0035] An acquisition module, configured to acquire the actual safety analysis data of the nuclear power plant and the simulated safety analysis data of the nuclear power simulation model under the same operation data, and the nuclear power simulation model is trained based on the operation data and the actual safety analysis data;

[0036] A determination module, configured to determine the status data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data;

[0037] An optimization module, configured to generate corresponding optimization strategies by combining the actual safety analysis data, the simulated safety analysis data, and the status data, and the optimization strategies are used to optimize the operation process of the nuclear power plant.

[0038] In a third aspect, an embodiment of the present disclosure provides an electronic device, including a memory, a processor, and a computer program stored on the memory and configured to run on the processor, and when the processor executes the computer program, it implements the optimization method for nuclear power plant operation according to any one of the first aspects.

[0039] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the optimization method for nuclear power plant operation according to any one of the first aspects.

[0040] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the optimization method for nuclear power plant operation according to any one of the first aspects.

[0041] On the basis of conforming to the common knowledge in the art, the above preferred conditions can be combined arbitrarily to obtain various preferred examples of the present disclosure.

[0042] The positive and progressive effects of the present disclosure are as follows: By using actual safety analysis data and simulation analysis data, an optimization strategy for the operation of a nuclear power plant can be generated, the dynamic working conditions of the nuclear power plant can be accurately restored, and real-time comparison and feedback with the actual safety analysis data are carried out to ensure the accuracy and real-time performance of the nuclear power simulation model. The generation of the optimization strategy can improve the operation efficiency and safety. Precise working condition simulation and optimization decision-making can be carried out throughout the life cycle of the nuclear power plant, the emergency response ability of the staff can be improved, the energy utilization can be optimized, the failure risk can be reduced, more intelligent and automated management of the nuclear power plant can be realized, and the safe operation of the nuclear power plant can be effectively guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A schematic diagram of an application environment provided by an exemplary embodiment of the present disclosure;

[0044] Figure 2 A first flowchart of an optimization method for the operation of a nuclear power plant provided by an exemplary embodiment of the present disclosure;

[0045] Figure 3 A second flowchart of an optimization method for the operation of a nuclear power plant provided by an exemplary embodiment of the present disclosure;

[0046] Figure 4 A third flowchart of an optimization method for the operation of a nuclear power plant provided by an exemplary embodiment of the present disclosure;

[0047] Figure 5 A module diagram of an optimization system for the operation of a nuclear power plant provided by an exemplary embodiment of the present disclosure;

[0048] Figure 6 A structural diagram of an electronic device provided by an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] The present disclosure will be further described below by way of examples, but the present disclosure is not limited to the scope of the described examples.

[0050] In the embodiments of the present disclosure, prefix words such as "first" and "second" are only used to distinguish different described objects, and have no restrictive effect on the position, order, priority, quantity, content, etc. of the described objects. The use of prefix words such as ordinal numbers for distinguishing described objects in the embodiments of the present disclosure does not constitute a restriction on the described objects. The statements of the described objects shall refer to the descriptions in the claims or the context of the embodiments, and no redundant restrictions shall be formed due to the use of such prefix words. In addition, in the description of this embodiment, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0051] In the embodiments of the present disclosure, the processing of the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information involved all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0052] Please refer to Figure 1 , Figure 1 which is a schematic diagram of an application environment provided by an embodiment of this application. The schematic diagram includes a terminal 100 and a server 110. Among them, a nuclear power simulation model is loaded on the server 110. After receiving the operation data and actual safety analysis data uploaded by the terminal 100, the server 110 can train the nuclear power simulation model through the operation data and actual safety analysis data. In addition, the server 110 also determines the state data of the nuclear power plant according to the actual safety analysis data and the simulated safety analysis data, and generates an operation strategy for the nuclear power plant in combination with these data.

[0053] Specifically, the server 110 obtains the actual safety analysis data of the nuclear power plant and the simulated safety analysis data of the nuclear power simulation model under the same operation data. The nuclear power simulation model is trained based on the operation data and the actual safety analysis data; determines the state data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data; combines the actual safety analysis data, the simulated safety analysis data, and the state data to generate a corresponding optimization strategy, and the optimization strategy is used to optimize the operation process of the nuclear power plant.

[0054] In the embodiments of the present disclosure, Figure 1 the terminal 100 shown may be an entity device of types such as a desktop computer, a tablet computer, a notebook computer, a digital assistant, a smart wearable device, etc.; among them, the smart wearable device may include a smart bracelet, a smart watch, smart glasses, a smart helmet, etc. Of course, the terminal 100 is not limited to the above-mentioned electronic devices with a certain entity, and it may also be software running on the above-mentioned electronic devices. For example, the terminal 100 may be a web page or an application.

[0055] Optionally, the terminal 100 may include a display screen, a storage device, and a processor connected by a data bus. Among them, the display screen can be used for actual safety analysis data, simulated safety analysis data, status data, optimization strategies, etc. The display screen can be a touch screen of a mobile phone or a tablet computer, etc. The storage device is used to store actual safety analysis data, simulated safety analysis data, status data, optimization strategies, or other data materials, etc. The storage device can be the memory of the terminal 100, or a storage device such as a smart media card, a secure digital card, or a flash card. The processor can be a single-core or multi-core processor.

[0056] Optionally, the terminal 100 can be set up at the workstation of the nuclear power plant management personnel, or can be set up in the nuclear power plant. The actual safety analysis data under the same operating data sent by the terminal 100 can be obtained through sensors inside the nuclear power plant connected by communication.

[0057] In the embodiments of the present application, the server can be, for example, Figure 1 the server 110 shown in the figure, or other computer terminals with the same functions as the server, or similar computing devices. Further, the server 110 can be replaced by a server system, an operation platform, or a server cluster including multiple servers.

[0058] For example, the server cluster includes multiple servers, and each server can undertake different steps in the entire solution. For example, the first server obtains the actual safety analysis data of the nuclear power plant under the same operating data sent by the terminal. The second server obtains simulated safety analysis data based on the nuclear power simulation model trained based on the operating data and the actual safety analysis data. The third server determines the status data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data. The fourth server combines the actual safety analysis data, the simulated safety analysis data, and the status data to generate an optimization strategy for optimizing the operation process of the nuclear power plant.

[0059] The server that loads the nuclear power simulation model realizes cloud distributed computing, and through cloud computing, it elastically schedules computing power resources for high-precision multi-physics field coupling simulation. At the same time, the server allows multiple terminal staff and engineers to remotely access simultaneously to carry out joint emergency drills or design verification.

[0060] Each server in the above-mentioned server cluster can establish a connection relationship through a wireless link, and can also establish a connection relationship through a wired link. Optionally, each server can be placed in the same computer room, or can be placed in different computer rooms.

[0061] The following introduces an optimization method for the operation of a nuclear power plant provided by an embodiment of the present disclosure. Figure 2 It is a schematic flowchart of an optimization method for the operation of a nuclear power plant provided by an embodiment of the present disclosure. This specification provides method operation steps such as in the embodiment or flowchart, but based on routine or non-creative labor, more or fewer operation steps may be included. The step order listed in the embodiment is only one way among the execution orders of numerous steps and does not represent the only execution order. When the actual system or server product executes, it can be executed in the order shown in the embodiment or the drawing or executed in parallel (for example, in an environment of parallel processors or multi-threaded processing). Specifically, as Figure 2 shown, the optimization method includes:

[0062] S201. Obtain the actual safety analysis data of the nuclear power plant and the simulated safety analysis data of the nuclear power simulation model under the same operating data.

[0063] Among them, the nuclear power simulation model is trained based on the operating data and the actual safety analysis data. During the operation of the nuclear power plant, various different working conditions will occur, such as normal operation conditions, anticipated operational occurrences, accident conditions, etc. The normal operation condition means that the reactor is in a critical state and the unit is connected to the grid and operates at rated power or low power; the anticipated operational occurrence means an event that may occur during normal operation; the accident condition includes design basis accidents and severe accidents. The change of the operating data during the operation of the nuclear power plant can reflect the actual working conditions of the operation process of the nuclear power plant. For example, whether the nuclear power plant is in a normal condition or an accident condition, and the analysis of the operating status of the nuclear power plant usually takes the safety analysis data of the nuclear power plant as a benchmark.

[0064] Therefore, in this embodiment, it is necessary to determine the safety analysis data of the nuclear power plant. Among them, some safety analysis data can be directly obtained through the sensors set inside the nuclear power plant, while some safety analysis data cannot be directly obtained. Therefore, in this embodiment, two types of actual safety analysis data and simulated safety analysis data are set. The actual safety analysis data is the data that can be directly obtained through the sensors set inside the nuclear power plant, and there is partial or complete overlap between the actual safety analysis data and the simulated safety analysis data.

[0065] In addition, during the process of collecting the actual safety analysis data, since many reactions in the nuclear power plant occur inside the subsystems, general collection methods are difficult to directly obtain all the safety analysis data. Therefore, in this embodiment, it is necessary to obtain simulated simulation data through the nuclear power simulation model for supplementation. Since the nuclear power simulation model is applicable to full-condition simulation of the nuclear power plant, the nuclear power simulation model can accurately reflect the actual situation of each subsystem in the nuclear power plant, and thus can also obtain the simulated operation data reflecting the inside of each subsystem.

[0066] Specifically, the operation data can be classified into multiple types, such as initial state parameters, system configuration and boundary conditions, accident types, staff intervention parameters, and environmental and physical model parameters. The initial state parameters include reactor power, initial pressure, initial coolant flow rate, initial core power distribution: axial and radial distribution tables; the system configuration and boundary conditions include the status of safety systems (such as whether the high-pressure injection system is available), the status of various valves (such as whether the safety valve and exhaust valve are open), and external cold source conditions (such as whether the external cooling water supply is normal); the accident types include large break loss of coolant accident (LOCA), station blackout (SBO), loss of secondary side heat sink (LORHR), accident start time and trigger conditions; the staff intervention parameters include: injection start time and flow rate, exhaust valve opening time, whether the containment spray is enabled; the environmental and physical model parameters include thermal conductivity, specific heat capacity, material properties, chemical reaction model parameters (such as zirconium-water reaction rate), and radiation heat transfer model options.

[0067] The safety analysis data can be classified into multiple types, such as thermal-hydraulics, core status, containment behavior, and radioactive release. The thermal-hydraulic parameters include the change of core outlet temperature over time (such as rising from 1200°C to 2700°C), coolant pressure (such as the pressure dropping to 2 MPa), and the change of coolant inventory (such as complete loss within a few hours); the core degradation behavior includes the fuel melting ratio (such as the melting rate reaching 80%), the loss of control rods, and the core relocation time (such as dropping to the bottom of the pressure vessel after 2.5 hours); the behavior of the pressure vessel and containment includes the internal pressure and temperature of the pressure vessel (used to judge whether there is melting through), the change of containment pressure, and the hydrogen volume concentration in the containment; the radioactive release information includes the release amount of active gases, the release amounts of key nuclides such as iodine and cesium, and the leakage path (such as discharged through the main steam pipeline); the environmental impact parameters include the dose rate time series curve and the estimated dose of air / ground deposition outside the perimeter; the severe accident management instructions include the recommended core injection window and the timing of hydrogen discharge / control.

[0068] It should be noted that the operation data and safety analysis data are not limited to the above examples and can be specifically selected according to the actual situation.

[0069] In one embodiment, regarding the nuclear power simulation model, specifically:

[0070] The nuclear power simulation model can achieve full coverage of operating conditions and complete the simulation of the entire life cycle process of a nuclear power plant, from the initial cold start, through steady-state operation at different power levels, to the final hot shutdown. During the startup phase, the nuclear power simulation model can accurately simulate the gradual increase of the neutron flux in the reactor core, the parameter changes in the primary and secondary loop systems, and the coordinated startup process of various auxiliary systems, providing detailed startup guidance and risk judgment for the staff. During normal operation, it can continuously simulate the operating states of various systems in the nuclear power plant under different loads, including conventional power regulation, regular equipment switching and other operation scenarios, ensuring that the nuclear power plant always operates in an efficient and stable state. During the shutdown process, the nuclear power simulation model can meticulously simulate the process of core cooling, residual heat removal, and the gradual depressurization and cooling of the system, helping the staff to formulate a safe and reliable shutdown plan.

[0071] The nuclear power simulation model adopts a cloud-based multi-model adaptive simulation framework (C-MMASF), including a coupled thermodynamics and fluid dynamics model, a neutron kinetics model, and an adaptive computing and resource scheduling model.

[0072] The coupled thermodynamics and fluid dynamics model combines the finite element method and the finite volume method to simulate the thermal-fluid coupling effect.

[0073] The neutron kinetics model describes the diffusion and absorption processes of neutrons in the reactor and uses the diffusion equation to simulate the neutron flux. The neutron diffusion equation is: . In the formula, is the neutron flux, is the diffusion coefficient, is the absorption cross section, is the neutron multiplication factor, is the fission cross section.

[0074] The adaptive computing and resource scheduling model dynamically selects the simulation accuracy and allocates computing resources according to resource scheduling and adaptive accuracy control.

[0075] The resource scheduling formula is , where N is the total number of computing nodes, is the computing resource consumption (time and power consumption) of the i-th node, is the scheduling decision variable (0 or 1, 0 means not allocated, 1 means allocated).

[0076] The adaptive accuracy control formula is , where M is the number of model complexity levels, is the computing accuracy factor (the value range is between 0 and 1, but not limited to this), is the current model error, To calculate the cost factor (the value range is 0 to 10, but not limited to this). In this model, the combination of the calculation accuracy factor and the calculation cost factor effectively controls the dynamic balance between the simulation accuracy and the computing resources. As the simulation complexity changes, the model will dynamically adjust these two factors according to the error evaluation.

[0077] S202. Determine the status data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data.

[0078] Among them, the status data may include fault data reflecting the failure of the neutron subsystem of the nuclear power plant or the change of a certain key safety analysis data in the nuclear power plant within a preset time period, etc. These data are all used to reflect the current or future state of the nuclear power plant during operation, so as to generate more accurate optimization strategies for the operation of the nuclear power plant in the later optimization process.

[0079] S203. Combine the actual safety analysis data, the simulated safety analysis data and the status data to generate corresponding optimization strategies.

[0080] Among them, the optimization strategy is used to optimize the operation process of the nuclear power plant, covering aspects such as load dispatching, equipment maintenance, and emergency response of the nuclear power plant. As an example, the optimization strategy can be generated by a large language model through learning on an expert knowledge base. A large number of optimization strategies corresponding one-to-one to the safety analysis data and / or status data are stored in the expert knowledge base. By searching for and combining the optimization strategies matching the safety analysis data and / or status data in the expert knowledge base, the required optimization strategy in this embodiment can be obtained. In addition, the generation scheme of the optimization strategy in this embodiment is not limited to this and can be selected according to the actual situation.

[0081] Specifically, the load dispatching optimization can automatically generate a load dispatching plan for the nuclear power plant based on the real-time power generation status and the power grid demand prediction. By analyzing the operation status of the equipment, the system can optimize the load distribution, ensure the efficient and stable operation of each unit, and avoid the instability caused by load fluctuations.

[0082] The equipment maintenance and resource scheduling optimization can automatically generate an equipment maintenance and resource scheduling plan according to the operation status, fault prediction and maintenance requirements of the equipment, ensuring that all maintenance work of the nuclear power plant can be carried out in a timely and effective manner, and avoiding the impact of equipment failures on the overall operation.

[0083] The emergency response and operation support can, when the nuclear power plant encounters a sudden failure or abnormal situation, the optimization decision-making module can provide detailed emergency operation suggestions according to the emergency response plan of the simulation, and automatically adjust the system control parameters, helping the staff to take correct emergency measures in the shortest time and reducing the losses of the accident.

[0084] In one embodiment, the optimization method further includes a training step of a nuclear power simulation model. Refer to Figure 3 , and the training step includes:

[0085] S301. Input the operation data into a pre-constructed nuclear power simulation model.

[0086] Among them, the pre-constructed nuclear power simulation model can refer to the content exemplified in the above embodiment.

[0087] S302. Construct an objective function based on the output of the nuclear power simulation model and the actual safety analysis data.

[0088] Setting the objective function includes:

[0089]

[0090] Among them, is the output of the nuclear power simulation model, is the actual safety analysis data, N is the total number of sample points, is the target parameter of the simulation model, is the input operation data.

[0091] S303. With the goal of minimizing the objective function, adjust the parameters of the nuclear power simulation model through a hybrid optimization algorithm to obtain a trained nuclear power simulation model.

[0092] Among them, the hybrid optimization algorithm is constructed based on the particle swarm optimization algorithm and the neural network algorithm. The hybrid optimization algorithm combines the advantages of the particle swarm algorithm (PSO) and the neural network algorithm (DNN), and can dynamically correct the nuclear power simulation model and parameters.

[0093] In a specific embodiment,

[0094] Refer to Figure 4 , and step S303 includes:

[0095] S3031. Generate an initial population. Among them, the initial population includes multiple particles, and one particle corresponds to a candidate solution set of parameters.

[0096] S3032. In each round of iteration process, calculate the fitness of each particle in the population in turn, update the best particle in the population, and update the position and velocity of each particle in the population until the sum of the fitness of all particles in the population is less than a preset threshold.

[0097] The velocity update formula is:

[0098] ;

[0099] The position update formula is:

[0100] ;

[0101] Among them, w is the inertia weight, which is used to control the influence of the current velocity of the particle on the next velocity update and affects the balance of the search process. is the velocity of particle i in the k-th iteration. and are learning factors, which are used to control the degree of attraction of the particle to its own optimal solution and the global optimal solution. and are random numbers uniformly distributed in the range of , which is used to increase randomness. is the position of particle i in the k-th iteration, representing the parameter set of the nuclear power simulation model. is the optimal solution of particle i itself (the parameter set corresponding to the minimum of the objective function), is the global optimal solution, representing the optimal solution among all particles.

[0102] S3033. In each round of iteration, a loss function is constructed based on the output of the nuclear power simulation model and the actual safety analysis data, and the output of the nuclear power simulation model is adjusted by the neural network algorithm with the goal of minimizing the gradient of the loss function.

[0103] Specifically, the neural network algorithm is used to correct the output of the nuclear power simulation model in real time to make it closer to the actual safety analysis data. The loss function formula of the neural network algorithm is:

[0104]

[0105] Among them, are neural network parameters, is the output prediction value, is the actual measured value of the nuclear power plant.

[0106] Neural network parameters are updated according to the gradient of the loss function in each iteration through the backpropagation method. That is . Among them, are the parameters (weights and biases) of the current deep neural network at the k-th iteration, is the learning rate.

[0107] ]>S3034. Take the best positions of each particle in the population as the global optimal solution of the parameters to obtain the trained nuclear power simulation model.

[0108] Specifically, after the particles are updated in each round of iteration, it is necessary to calculate and update the global optimal solution , is the parameter set with the minimum objective function among all particles.

[0109] In one embodiment, it is possible to comprehensively perceive and real-time monitor the operation process of a nuclear power plant, achieve a comprehensive analysis of the dynamic situation of the nuclear power plant, and the status data includes the change data of the safety analysis data of the nuclear power plant within a preset time period. Step S202 specifically includes:

[0110] Determine the change data of the safety analysis data of the nuclear power plant within a preset time period based on the actual safety analysis data and the simulated safety analysis data.

[0111] Specifically, through status data, actual safety analysis data, simulated safety data, etc., it is possible to comprehensively perceive the operation status of on-site equipment in the nuclear power plant. By summarizing, analyzing, and processing the above data, it is possible to predict the change data of the safety analysis data of the nuclear power plant within a preset time period, and the change data specifically includes the future change trajectory. Among them, the safety analysis data for future prediction can be partial key safety analysis data or all safety analysis data.

[0112] In addition, the generation of the change data can be obtained by learning the historical safety analysis data through a deep learning model and the above nuclear power simulation model, which will not be elaborated here and can be specifically selected according to the actual situation.

[0113] On this basis, it is also possible to generate an optimization strategy that matches the change data. Step S203 specifically:

[0114] Combine the actual safety analysis data, simulated safety analysis data, and change data to generate a corresponding optimization strategy.

[0115] Among them, the generation of the optimization strategy depends not only on the actual safety analysis data but also on the change data, and can be targeted for optimization of the changes that will occur in the nuclear power plant in the future.

[0116] In one embodiment, it is also possible to timely identify fault data such as equipment failures, operation anomalies, and other potential problems in the nuclear power plant through the comparison and analysis of the simulated safety analysis data and the actual safety analysis data of the nuclear power plant, and generate a corresponding optimization strategy for it. Specifically, the status data also includes fault data. Step S202 specifically includes:

[0117] Determine the fault data based on the actual safety analysis data, simulated safety analysis data, and change data.

[0118] Among them, the fault data includes the fault content and the fault cause. The fault content indicates the specific fault situation, such as changes caused by temperature and pressure, etc., while the fault cause indicates the fault situation caused by specific equipment.

[0119] In a specific embodiment, the fault content is determined based on actual safety analysis data, simulated safety analysis data, and change data. Taking the fault content as the target event and at least one of the actual safety analysis data, simulated safety analysis data, and change data as the state, the dynamic fault tree is used for analysis to determine the fault cause.

[0120] The dynamic fault tree is a dynamic fault tree based on a time series probability graph model (TSPG-DFT). The dynamic fault tree integrates time series correlation analysis and multi-modal data, and its mathematical model construction includes a state transition equation and joint probability calculation.

[0121] First, the hidden Markov model (HMM) is used to describe the system state evolution:

[0122]

[0123] State transition matrix: , represents the probability of transitioning from state i to state j, and N is the number of system states.

[0124] Observation probability matrix: , represents the probability of observing k in state j, and K is the number of observation values.

[0125] Initial state distribution: , represents the probability that the system is in state i at the initial moment.

[0126] Secondly, joint probability calculation is performed:

[0127]

[0128] where M is the total number of fault paths, T is the total length of the time series, is the conditional probability that event occurs at time t given that the state of event is known at time . represents the conditional probability of observing given that the system state is known at time t, is the observation sequence at time t.

[0129] In a specific embodiment, future fault data reflecting equipment failures in a nuclear power plant can also be identified through learning of equipment health data and fault data, and predictions can be made in advance. For example, by analyzing the data of power transformers, possible loss conditions or fault points of the transformers can be identified in advance, and early warnings can be issued to prompt the management to perform necessary repairs or replacements.

[0130] In this embodiment, various possible faults can be identified through actual safety analysis data and simulated safety analysis data. For example, if the temperature of the reactor rises abnormally, the system will quickly determine whether it is a problem with the temperature control system or a failure of the cooling system by combining the equipment health status and simulation prediction, and automatically locate the problem. Once a fault is detected, based on dynamic fault tree analysis, detailed fault causes can be provided to help the nuclear power plant quickly take repair measures.

[0131] In one embodiment, step S203 specifically includes:

[0132] Combine the actual safety analysis data, simulated safety analysis data, and fault data to generate corresponding optimization strategies.

[0133] In this embodiment, comprehensive optimization strategies can be provided for the nuclear power plant based on the actual safety analysis data, simulated safety analysis data, and fault data.

[0134] In one embodiment, before obtaining the actual safety analysis data of the operation data in step S201, since the collected raw data often contains noise, missing values, or abnormal data, the actual safety analysis data also needs to be preprocessed, specifically including data cleaning detection and default value processing. Through means such as data cleaning, denoising, and filtering, ensure that the data input into the simulation system meets the model requirements and avoid misleading the simulation results. The module will also mark or alarm the unstable or data deviating from the normal range to ensure data quality.

[0135] Data cleaning detection includes: (1) Multi-dimensional detection: Combine statistical analysis (such as the 3σ rule) and machine learning algorithms (such as isolation forest) to identify abnormal data points outside the normal range; (2) Statistical detection: Calculate the mean and standard deviation of the data, and mark the points deviating more than 3 times the standard deviation from the mean; (3) Process rule constraints: According to the nuclear power operation specifications (such as the pressure of the pressurizer must be 15.5 ± 0.2 MPa), filter the data that obviously violates the physical laws.

[0136] The processing of the detected outliers includes: (1) A small number of outliers are directly deleted and supplemented with the average value of adjacent data points; (2) Continuous outliers are filled by interpolation of adjacent data.

[0137] The processing of default values includes: (1) In steady-state conditions, cubic spline interpolation or polynomial fitting is used to smoothly fill the data using adjacent time points; (2) In transient conditions, extrapolation is performed based on the prediction results of simulation data to ensure that the filled values conform to physical laws; (3) For sensor data with spatial distribution (such as core temperature measurement points), Kriging interpolation is used to fill in combination with the spatial correlation of surrounding measurement points.

[0138] An exemplary embodiment of the present disclosure provides an optimization system for nuclear power plant operation. Refer to Figure 5 , the optimization system includes:

[0139] An acquisition module 51, configured to acquire the actual safety analysis data of the nuclear power plant and the simulated safety analysis data of the nuclear power simulation model under the same operation data, where the nuclear power simulation model is trained based on the operation data and the actual safety analysis data;

[0140] A determination module 52, configured to determine the state data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data;

[0141] An optimization module 53, configured to generate a corresponding optimization strategy by combining the actual safety analysis data, the simulated safety analysis data, and the state data, where the optimization strategy is used to optimize the operation process of the nuclear power plant.

[0142] In one embodiment, the optimization system further includes a training module for the nuclear power simulation model;

[0143] The training module includes:

[0144] A first construction unit, configured to input the operation data into a pre-constructed nuclear power simulation model;

[0145] A second construction unit constructs an objective function based on the output of the nuclear power simulation model and the actual safety analysis data;

[0146] An adjustment unit, configured to adjust the parameters of the nuclear power simulation model by a hybrid optimization algorithm with the goal of minimizing the objective function, so as to obtain a trained nuclear power simulation model, where the hybrid optimization algorithm is constructed based on the particle swarm optimization algorithm and the neural network algorithm.

[0147] In one embodiment, the adjustment unit is further configured to:

[0148] Generate an initial population, where the initial population includes multiple particles, and one particle corresponds to a candidate solution set of parameters;

[0149] In each round of iteration, calculate the fitness of each particle in the population in turn, update the best particle in the population, and update the position and velocity of each particle in the population until the sum of the fitnesses of all particles in the population is less than a preset threshold;

[0150] In each round of iteration, construct a loss function based on the output of the nuclear power simulation model and the actual safety analysis data, and adjust the output of the nuclear power simulation model by the neural network algorithm with the goal of minimizing the gradient of the loss function;

[0151] Take the best position of each particle in the population as the global optimal solution of the parameters to obtain a trained nuclear power simulation model.

[0152] In one embodiment, the status data includes the change data of the safety analysis data of the nuclear power plant within a preset time period;

[0153] The determination module 52 is further configured to determine the change data of the safety analysis data of the nuclear power plant within a preset time period based on the actual safety analysis data and the simulated safety analysis data;

[0154] The optimization module 53 is further configured to generate a corresponding optimization strategy by combining the actual safety analysis data, the simulated safety analysis data, and the change data.

[0155] In one embodiment, the status data further includes fault data;

[0156] The determination module 52 is further configured to determine the fault data based on the actual safety analysis data, the simulated safety analysis data, and the change data;

[0157] The optimization module 53 is further configured to generate a corresponding optimization strategy by combining the actual safety analysis data, the simulated safety analysis data, and the fault data.

[0158] In one embodiment, the fault data includes the fault content and the fault cause;

[0159] The determination module 52 is further configured to:

[0160] Determine the fault content based on the actual safety analysis data, the simulated safety analysis data, and the change data;

[0161] Taking the fault content as the target event and at least one of the actual safety analysis data, the simulated safety analysis data, and the change data as the state, analyze through a dynamic fault tree to determine the fault cause.

[0162] For the system embodiment, since it basically corresponds to the method embodiment, the relevant parts can be referred to the partial description of the method embodiment. The system embodiments described above are only illustrative, where the units described as separate components may or may not be physically separated, and the components as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution.

[0163] An exemplary embodiment of the present disclosure further provides an electronic device. Refer to Figure 6 , the electronic device includes a memory, a processor, and a computer program stored on the memory and configured to run on the processor. When the processor executes the computer program, it implements the optimization method for nuclear power plant operation described in any of the above embodiments. Figure 6The illustrated electronic device 60 is merely an example and should not impose any limitation on the functions and usage scope of the embodiments of the present disclosure.

[0164] As Figure 6 shown, the electronic device 60 may be presented in the form of a general-purpose computing device. For example, it may be a server device. The components of the electronic device 60 may include, but are not limited to: at least one of the above-mentioned processors 61, at least one of the above-mentioned memories 62, and a bus 63 that connects different system components (including the memory 62 and the processor 61).

[0165] The bus 63 includes a data bus, an address bus, and a control bus.

[0166] The memory 62 may include volatile memory, such as a random access memory (RAM) 621 and / or a cache memory 622, and may further include a read-only memory (ROM) 623.

[0167] The memory 62 may also include a program tool 625 (or utility) having a set (at least one) of program modules 624. Such program modules 624 include, but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include the implementation of a network environment.

[0168] The processor 61 executes various functional applications and data processing by running computer programs stored in the memory 62, such as the optimization method for nuclear power plant operation provided in any of the above embodiments.

[0169] The electronic device 60 may also communicate with one or more external devices 64 (such as a keyboard, a pointing device, etc.). Such communication may be performed through an input / output (I / O) interface 65. And, the electronic device 60 may also communicate with one or more networks (such as a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 66. As shown in the figure, the network adapter 66 communicates with other modules of the electronic device 60 through the bus 63. It should be understood that although not shown in the figure, other hardware and / or software modules may be used in combination with the electronic device 60, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (redundant array of independent disks) systems, tape drives, and data backup storage systems, etc.

[0170] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / modules. Conversely, the features and functions of one unit / modules described above can be further divided and embodied by multiple units / modules.

[0171] Embodiments of the present disclosure also provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the optimization method for nuclear power plant operation provided in any of the above embodiments.

[0172] Among them, the computer-readable storage medium can be more specifically but not limited to: portable disks, hard disks, random access memories, read-only memories, erasable programmable read-only memories, optical storage devices, magnetic storage devices, or any suitable combination of the above.

[0173] Embodiments of the present disclosure also provide a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the optimization method for nuclear power plant operation described in any one of the above.

[0174] Among them, the program code for executing the computer program product of the present disclosure can be written in any combination of one or more programming languages, and the program code can be executed completely on the user device, partially on the user device, executed as an independent software package, partially on the user device and partially on a remote device, or executed completely on a remote device.

[0175] Although the specific embodiments of the present disclosure are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present disclosure is defined by the appended claims. Without departing from the principles and essence of the present disclosure, those skilled in the art can make various changes or modifications to these embodiments, but these changes and modifications all fall within the protection scope of the present disclosure.

Claims

1. An optimization method for the operation of a nuclear power plant, characterized in that, The optimization method includes: Obtaining the actual safety analysis data of a nuclear power plant and the simulated safety analysis data of a nuclear power simulation model under the same operating data, where the nuclear power simulation model is obtained by performing a full-condition simulation based on the nuclear power plant; Determining the status data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data; Combining the actual safety analysis data, the simulated safety analysis data, and the status data to generate a corresponding optimization strategy, where the optimization strategy is used to optimize the operation process of the nuclear power plant.

2. The optimization method according to claim 1, wherein The optimization method further includes a training step for the nuclear power simulation model; The training step includes: Inputting the operating data into a pre-constructed nuclear power simulation model; Constructing an objective function based on the output of the nuclear power simulation model and the actual safety analysis data; Taking the minimum of the objective function as the goal, adjusting the parameters of the nuclear power simulation model through a hybrid optimization algorithm, and obtaining the trained nuclear power simulation model, where the hybrid optimization algorithm is constructed based on a particle swarm optimization algorithm and a neural network algorithm.

3. The optimization method according to claim 2, wherein Taking the minimum of the objective function as the goal, adjusting the parameters of the nuclear power simulation model through a hybrid optimization algorithm, and obtaining the trained nuclear power simulation model, including: Generating an initial population, where the initial population includes multiple particles, and one particle corresponds to a candidate solution set of the parameters; In each round of iteration, successively calculating the fitness of each particle in the population, updating the best particle in the population, and updating the position and velocity of each particle in the population until the sum of the fitnesses of all particles in the population is less than a preset threshold; In each round of iteration, constructing a loss function based on the output of the nuclear power simulation model and the actual safety analysis data, and adjusting the output of the nuclear power simulation model through a neural network algorithm with the goal of minimizing the gradient of the loss function; Taking the best positions of the particles in the population as the global optimal solution of the parameters, and obtaining the trained nuclear power simulation model.

4. The optimization method according to claim 1, characterized in that The status data includes the change data of the safety analysis data of the nuclear power plant within a preset time period; The determining the status data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data includes: Determining the change data of the safety analysis data within a preset time period based on the actual safety analysis data and the simulated safety analysis data; The combining the actual safety analysis data, the simulated safety analysis data, and the status data to generate a corresponding optimization strategy includes: Combining the actual safety analysis data, the simulated safety analysis data, and the change data to generate a corresponding optimization strategy.

5. The optimization method according to claim 4, characterized in that, The status data further includes fault data; After determining the change data of the safety analysis data of the nuclear power plant within a preset time period based on the actual safety analysis data and the simulated safety analysis data, it includes: Determining the fault data based on the actual safety analysis data, the simulated safety analysis data, and the change data; The combining the actual safety analysis data, the simulated safety analysis data, and the status data to generate a corresponding optimization strategy includes: Generate corresponding optimization strategies by combining the actual safety analysis data, the simulated safety analysis data, and the fault data.

6. The optimization method according to claim 5, wherein The fault data includes fault content and fault cause. Determining the fault data based on the actual safety analysis data, the simulated safety analysis data, and the change data includes: Determining the fault content based on the actual safety analysis data, the simulated safety analysis data, and the change data; Taking the fault content as the target event and at least one of the actual safety analysis data, the simulated safety analysis data, and the change data as the state, and analyzing through a dynamic fault tree to determine the fault cause.

7. An optimization system for nuclear power plant operation, characterized in that, The optimization system includes: An acquisition module, configured to acquire the actual safety analysis data of a nuclear power plant and the simulated safety analysis data of a nuclear power simulation model under the same operating data, where the nuclear power simulation model is trained based on the operating data and the actual safety analysis data; A determination module, configured to determine the state data of the nuclear power plant based on the actual safety analysis data and the simulated safety analysis data; An optimization module, configured to generate corresponding optimization strategies by combining the actual safety analysis data, the simulated safety analysis data, and the state data, where the optimization strategies are used to optimize the operation process of the nuclear power plant.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and configured to run on the processor, characterized in that, When the processor executes the computer program, it implements the optimization method for the operation of a nuclear power plant according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the optimization method for the operation of a nuclear power plant according to any one of claims 1-6.

10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the optimization method for the operation of a nuclear power plant as described in any one of claims 1-6.

Images