Method for generating a nuclear power plant operating scheme, system and readable medium therefor

Abstract

The application provides a method for generating a nuclear power operation scheme, a system and a readable medium applicable to the method, and relates to the technical field of nuclear power. The method comprises the following steps: obtaining business data, step dependency relationship data, a trained business identification model and a plurality of trained step generation models; determining a business category corresponding to the business data according to the business data and the business identification model; selecting a step generation model from the plurality of step generation models as a current model according to the business category; determining a plurality of business steps corresponding to the business data according to the business data and the current model; and sorting the plurality of business steps according to the step dependency relationship data and a sorting algorithm, and determining the plurality of business steps after sorting as an operation scheme corresponding to the business data, wherein the step dependency relationship data is used to reflect the constraints of nuclear power plant rules and / or actual factors of the nuclear power plant on the business steps.

Description

Technical Field

[0001] This application relates primarily to the field of nuclear power technology, and in particular to a method for generating nuclear power operation schemes, as well as the applicable system and readable medium thereof. Background Technology

[0002] Risk control is crucial in nuclear power production; therefore, improving operational efficiency in production management is essential for enhancing nuclear power safety. In nuclear power production management, numerous business activities require the development of corresponding operational plans based on task objectives. These plans typically consist of multiple steps. The completeness, rationality, and execution sequence of these steps directly impact the effectiveness of implementation, the level of safety risk control, and production efficiency.

[0003] Currently, in some related technologies, operational plans mainly rely on experienced business personnel to manually develop them. As a result, these related technologies have the following problems: (1) Business experience is highly dependent on personal accumulation; (2) Similar business needs are repeatedly planned, resulting in a large workload for business personnel. Summary of the Invention

[0004] The technical problem to be solved by this application is to provide a method for generating nuclear power plant operation plans, as well as a suitable system and readable medium, which can automatically generate operation plans and ensure the reliability and effectiveness of the operation plans.

[0005] To address the aforementioned technical problems, this application provides a method for generating nuclear power operation schemes, applicable to nuclear power plants. The method includes: acquiring business data, step dependency data, a trained business identification model, and multiple trained step generation models. The business data includes the business object, business reason, business scope, and affected area corresponding to the nuclear power plant. Based on the business data and the business identification model, the method determines the business category corresponding to the business data. Based on the business category, it selects one step generation model from the multiple step generation models as the current model. Based on the business data and the current model, it determines multiple business steps corresponding to the business data. Based on the step dependency data and a sorting algorithm, it sorts the multiple business steps and determines the sorted multiple business steps as the operation scheme corresponding to the business data. The step dependency data reflects the constraints imposed on the business steps by nuclear power plant rules and / or actual factors of the nuclear power plant.

[0006] Optionally, the step dependency data includes a general dependency graph and multiple hierarchical dependency graphs. The steps of obtaining business data, step dependency data, a trained business identification model, and multiple trained step generation models further include: determining multiple business levels corresponding to the nuclear power plant; and constructing a general dependency graph and a hierarchical dependency graph corresponding to each business level based on multiple historical business data of the nuclear power plant and corresponding historical operation schemes.

[0007] Optionally, the steps of acquiring business data, step dependency data, a trained business identification model, and multiple trained step generation models further include: determining multiple business categories based on multiple business levels, multiple historical business data, and corresponding historical operation schemes; for each business category, training an initial generation model based on the historical business data and historical operation schemes corresponding to the business category to obtain a trained initial generation model; and for each business category, using the corresponding trained initial generation model as the step generation model corresponding to the business category.

[0008] Optionally, for each business category, the step of training an initial generation model based on the historical business data and historical operation plans corresponding to the business category to obtain a completed initial generation model further includes: standardizing the historical business steps in the historical operation plans according to the step dictionary to obtain standardized historical business steps, wherein the standardized historical business steps include action type, operation object, parameter constraints and risk level; and training an initial generation model based on the historical business data and standardized historical business steps corresponding to the business category to obtain a completed initial generation model.

[0009] Optionally, the step of sorting multiple business steps according to the step dependency data and the sorting algorithm, and determining the sorted business steps as the operation scheme corresponding to the business data, further includes: sorting multiple business steps according to the sorting algorithm to obtain sorted business steps; and performing sorting verification on the sorted business steps according to the general dependency graph and the hierarchical dependency graph corresponding to the business level where the business data is located to obtain sorted business steps.

[0010] Optionally, the step of determining the business category corresponding to the business data based on the business data and the business identification model further includes: encoding the business data to obtain an encoding vector; and inputting the encoding vector into the business identification model to obtain the business category corresponding to the business data.

[0011] Optionally, the calculation expression for the business category is: , In the formula For the business categories corresponding to the business data, Business Category The corresponding rating The length of the label sequence. The label transition matrix, For tags The business identification model outputs probabilities.

[0012] Optionally, the generation method further includes: obtaining the validity scores of the user execution plan and operation plan corresponding to the business data; determining the correction amount corresponding to the business data based on the user execution plan and the operation plan corresponding to the business data; constructing training samples based on the business data, the corresponding correction amount, the validity score, and the user execution plan; and retraining the corresponding business recognition model and / or step generation model based on the training samples.

[0013] To address the aforementioned technical problems, this application provides a nuclear power plant operation scheme generation system, comprising: a memory for storing instructions executable by a processor; and a processor for executing the instructions to implement the aforementioned nuclear power plant operation scheme generation method.

[0014] To address the aforementioned technical problems, this application provides a computer-readable medium storing computer program code, which, when executed by a processor, implements the aforementioned method for generating a nuclear power plant operation scheme.

[0015] Compared with existing technologies, this application has the following advantages: It effectively determines the business category of the business to be executed by business data and a business identification model, and processes the business data using a step generation model corresponding to that business category, thereby obtaining accurate multiple business steps. This effectively suppresses the problem of data interference caused by a single step generation model being used for a large number of business scenarios in nuclear power plants, which makes it difficult to effectively improve the generation accuracy of business steps related to certain business scenarios. Furthermore, multiple business steps are sorted using a sorting algorithm and step dependency data, ensuring that the final generated operation plan can be effectively applied to the business to be executed, and that the sorted business steps meet the rules and actual needs of the nuclear power plant during execution. Therefore, it not only automates the generation of operation plans for user reference, reducing the workload of business personnel, but also ensures the reliability and effectiveness of the operation plans. Attached Figure Description

[0016] The accompanying drawings are included to provide a further understanding of this application; they are incorporated into and constitute a part of this application. The drawings illustrate embodiments of this application and, together with this specification, serve to explain the principles of this application. In the drawings: Figure 1 This is a flowchart illustrating a method for generating a nuclear power operation scheme according to an embodiment of this application; Figure 2 yes Figure 1 A flowchart illustrating the sub-steps of step S1 in the illustrated embodiment; Figure 3 yes Figure 2 A flowchart illustrating the sub-step of step S14 in the illustrated embodiment; Figure 4yes Figure 1 A flowchart illustrating the sub-steps of step S2 in the illustrated embodiment; Figure 5 yes Figure 1 A flowchart illustrating the sub-steps of step S5 in the illustrated embodiment. Figure 6 yes Figure 1 The illustrated embodiment is a flowchart of the method for generating a nuclear power plant operation plan; Figure 7 This is a schematic diagram of a nuclear power operation scheme generation system according to an embodiment of this application. Detailed Implementation

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this application. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0018] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0019] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of this application. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following drawings denote similar items; therefore, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0020] Furthermore, although the terminology used in this application is selected from commonly known and used terms, some terms mentioned in this application's specification may have been chosen by the applicant according to his or her judgment, and their detailed meanings are explained in the relevant sections of the description herein. Moreover, this application is to be understood not only by the actual terms used, but also by the meaning implied by each term.

[0021] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more steps may be removed from these processes.

[0022] In some related technologies, nuclear power plants have accumulated a large number of historical success cases compiled by operational personnel based on procedures, historical cases, and personal experience over long-term operation. However, these historical success cases are usually stored in a scattered manner in the form of work orders, operation tickets, maintenance records, operation logs, and implementation reports. Furthermore, given the complex and numerous business scenarios within nuclear power plants, it is difficult to form automatically usable knowledge assets. In other related technologies, there are logical relationships between business steps, but some automatic step sequencing methods for operational plans cannot effectively identify these logical relationships or incorrectly confuse the logical relationships between different business scenarios, resulting in logical problems in step sequencing that prevent proper use.

[0023] In this regard, refer to Figure 1 One embodiment of this application proposes a method 100 for generating nuclear power plant operation plans (hereinafter referred to as generation method 100), and this generation method 100 is applied to a nuclear power plant. This generation method 100 is suitable for generating corresponding operation plans based on the operational data of the nuclear power plant. For example... Figure 1 As shown, generation method 100 includes the following steps.

[0024] Step S1 involves acquiring business data, step dependency data, a trained business identification model, and multiple trained step generation models. The business data includes the business object, business reason, business scope, and affected area corresponding to the nuclear power plant. In this embodiment, the business data corresponds to the business to be executed. The business to be executed can be comprehensively and accurately described through the business data to facilitate the subsequent generation of reliable and effective operational plans. The business data can include structured first business data and unstructured second business data associated with the business to be executed. The first business data can include information such as business sub-category, safety level, and regulatory department. The second business data includes the business object, business reason, business scope, and affected area. In this embodiment, the business object reflects the target corresponding to the business to be executed, the business reason reflects the problem or cause corresponding to the business to be executed, the business scope reflects the relevant equipment involved in the business to be executed, and the affected area reflects the area of ​​the nuclear power plant involved when the business to be executed is performed. For example, when the task to be performed is to overhaul the frequency converter of the turbine remote control system in the conventional island of a nuclear power plant, the task object is the frequency converter, the task reason is that the frequency converter's coil is overheating, the task scope is a portion of the circuits in the turbine remote control system, as well as temperature and pressure acquisition equipment, and the affected area is a specific room in the conventional island of the nuclear power plant. In this embodiment, a standardized task description template can be constructed based on the task object, task reason, task scope, and affected area, thereby enabling users to provide corresponding second task data based on the standardized task description template. Thus, facing different users' different expression habits and thinking methods, a relatively uniform second task data containing necessary information can be obtained through the standardized task description template, thereby enabling the generation of a more accurate operation plan based on the task data. It should be noted that this application does not limit the content of the second task data. In some embodiments, the second task data also includes risk factors, precautions, and expected goals. Risk factors reflect the risk type corresponding to the task to be performed, precautions reflect the user's corresponding needs for the task to be performed, and expected goals reflect the expected effect after the task to be performed is completed. For example, when the task to be performed is to overhaul the frequency converter of the remote control system for the steam turbine in the conventional island of a nuclear power plant, risk factors include high-voltage discharge and poor contact, and the expected goal is to restore the normal function of the frequency converter. In this embodiment, the first task data can be obtained by reading the corresponding field of the corresponding data in the existing system related to the task data, and the second task data can be obtained by manual input by the user.

[0025] Continue to refer to Figure 1 The step dependency data includes a general dependency graph and multi-level dependency graphs. This data reflects the dependencies between business steps. Further reference... Figure 2Step S1 includes the following sub-steps. Step S11 is to determine the multiple business levels corresponding to the nuclear power plant. In this embodiment, the multiple business levels may include the business site level, business unit level, system level, functional location level, and equipment category level. It can be understood that the business site level, business unit level, system level, functional location level, and equipment category level correspond to the plant level, unit level, system level, functional location level, and equipment level of the nuclear power plant, respectively, from largest to smallest. It should be noted that equipment with the same function in different business scenarios in a nuclear power plant can correspond to the same category. For example, the booster station in a nuclear power plant and the filter step-down transformer of a specific device in a nuclear power plant have great similarities in principle and operation, and thus both belong to the transformer category. However, in actual nuclear power business scenarios such as isolation operations, the specific business operations for the booster station and the filter step-down transformer are very different. Therefore, in this embodiment, based on the characteristics of nuclear power plants, the equipment and devices in nuclear power plants are divided into the above-mentioned multiple business levels, so that various equipment and devices with the same category but significantly different application scenarios can be more accurately classified, so as to generate accurate operation plans for the corresponding business data in the future.

[0026] Continue to refer to Figure 2 Step S12 involves constructing a general dependency graph and a hierarchical dependency graph for each business level based on multiple historical business data and corresponding historical operation plans of the nuclear power plant. In this embodiment, general dependency data and hierarchical dependency data for each business level of the nuclear power plant are extracted from historical business data and historical operation plans. A general dependency graph is then constructed based on the general dependency data, and corresponding hierarchical dependency graphs are constructed based on the hierarchical dependency data. This process standardizes and normalizes the rules, requirements, and actual factors related to the nuclear power plant, transforming them into identifiable and executable graph data, namely, the general dependency graph and multiple hierarchical dependency graphs. It is understandable that by collecting the business data and operation plans generated in real time during the operation of the nuclear power plant and using them as corresponding historical business data and historical operation plans, the actual needs and rules of the nuclear power plant can be continuously updated. This allows for reliable and timely updates and adjustments to the general dependency graph and hierarchical dependency graphs, thereby ensuring the reliability and effectiveness of subsequent operation plans generated based on the step dependency data. In some embodiments, the regulations and rules of the nuclear power plant can also be obtained, and the general dependency graph and the multi-level dependency graph can be further improved based on the regulations and rules.

[0027] Continue to refer to Figure 2Step S13 involves determining multiple business categories based on multiple business levels, historical business data, and corresponding historical operation plans. It should be noted that in nuclear power plants, to improve safety, specific operations are set according to the needs of different business scenarios. This results in a massive number of different types of business scenarios in nuclear power plants, with some scenarios exhibiting certain similarities. Therefore, in step S13, for each business level, multiple business categories are determined based on the historical business data and historical operation plans corresponding to that business level. It can be understood that business categories reflect the business scenario corresponding to the business to be executed. In this embodiment, determining business categories enables accurate classification of the business to be executed, thereby providing data support for generating operation plans for the subsequent business to be executed. It should be noted that this application does not limit whether the total number of business categories corresponding to each business level is equal. In some embodiments, the total number of business categories corresponding to each business level can be preset to ensure that the total number of business categories is the same, facilitating unified management. In some embodiments, a preset total number of business categories can be set for each business level, resulting in a different total number of business categories for each business level. This allows for a more reasonable business category classification based on the actual distribution of relevant business scenarios in the nuclear power plant. In some embodiments, classification reference features can be set, and then classification methods such as clustering algorithms can be used to automatically classify business categories based on historical business data and historical operation plans corresponding to each business level.

[0028] Continue to refer to Figure 2 Step S14 involves training an initial generative model for each business category based on its corresponding historical business data and operational procedures, resulting in a fully trained initial generative model. This initial generative model can be an existing deep learning model, such as the Transformer model, or it can be built by the developers themselves. Further reference... Figure 3 Step S14 includes the following sub-steps. Step S141 is to obtain standardized historical business steps based on the historical business steps in the standardized historical operation scheme of the step dictionary. The standardized historical business steps include action type, operation object, parameter constraints, and risk level. In this embodiment, the step dictionary may include step specifications corresponding to multiple different types of steps, and each step specification includes action type, operation object, parameter constraints, and risk level. For example, the expression of the step dictionary is: In the formula For step dictionary, For the first step in the dictionary The step specifications correspond to each type of business step. The corresponding expression for the step specification is: In the formula For the first The step specifications for each type of business step For the first The action types corresponding to each type of business step For the first The operation objects corresponding to each type of business step For the first Parameter constraints corresponding to each type of business step For the first The risk levels correspond to different types of business steps. In this embodiment, the action type, operation object, parameter constraints, and risk level can all be set with corresponding fixed selection ranges. This allows standardized historical business steps to provide easily understandable data features for the initial generation model when subsequently applied to its training, and ensures that the trained initial generation model has a standardized generated data format. In this embodiment, parameter constraints can be used to determine the limitations and other constraints of the corresponding operation actions. In this embodiment, the risk level can be used to determine the evaluation value corresponding to the action category and operation object. This evaluation value can be used for user reference to determine whether to adopt or adjust the business step. In this embodiment, the risk levels corresponding to different business steps can be preset based on human experience.

[0029] Continue to refer to Figure 3 Step S142 involves training an initial generation model for each business category based on the corresponding historical business data and standardized historical business steps, resulting in a completed initial generation model. In this embodiment, standardized historical business steps are used as part of the training data. This ensures that the completed initial generation model effectively learns key features and generates data in a standardized format. In other words, the business steps generated by the completed initial generation model also include information such as action type, operation object, parameter constraints, and risk level, thus providing users with effective and necessary reference data.

[0030] Continue to refer to Figure 2 Step S15 involves using the corresponding trained initial generative model as the step-generation model for each business category. It's understandable that training the initial generative model with targeted historical business data and operational procedures for each business category, thereby generating the corresponding step-generation model, improves the accuracy and reliability of subsequent generation of operational procedures for each business category. Furthermore, since nuclear power plants accumulate a large amount of historical data available for training, classifying this data by business category effectively categorizes this large amount of trainable data, thus improving the training efficiency of the step-generation model.

[0031] Continue to refer to Figure 1 Step S2 involves determining the business category corresponding to the business data based on the business data and the business identification model. In this embodiment, training data can be constructed based on historical business data and corresponding business categories, and this training data can be used to train an existing large model or neural network model to obtain the business identification model. Further reference... Figure 4 Step S2 includes the following steps. Step S21 is to encode the business data to obtain an encoding vector. In this embodiment, the structured first business data can be directly encoded according to classification, and the unstructured second business data expressed in natural language can be encoded using the BERT model. Subsequently, the encoded first and second business data can be integrated to obtain the encoding vector corresponding to the business data. It should be noted that in the relevant systems of nuclear power plants, the structured first business data has a corresponding encoding. For example, each specific category in the business sub-category has a corresponding letter string for representation, where the letter string can include multiple sequentially arranged letters. Therefore, the encoding of this specific category can be the corresponding letter string.

[0032] Continue to refer to Figure 4 Step S22 involves inputting the encoded vector into the service identification model to obtain the service category corresponding to the service data. In this embodiment, the service identification model includes a BiLSTM-CRF model. The BiLSTM-CRF model combines a bidirectional long short-term memory network and a conditional random field, thereby fully utilizing the contextual information in the encoded vector to improve prediction accuracy. The expression for the forward propagation process of the bidirectional long short-term memory network is as follows: In the formula For time steps The hidden state, For time steps The input feature vector, For time steps The hidden state, This is the state update function for the forward Long Short-Term Memory (LSTM) network unit. The expression for the backpropagation process of the bidirectional LSM network is: In the formula For the reverse long short-term memory network at time step The hidden state, For the reverse long short-term memory network at time step The hidden state, This is the state update function for the inverse long short-term memory network unit. Therefore, the bidirectional long short-term memory network updates at time steps... The final hidden state is the output. The expression is: In this embodiment, the calculation expression for the business category is: , , In the formula For the business categories corresponding to the business data, Business Category The corresponding rating The length of the label sequence. The label transition matrix, For tags The business identification model outputs probabilities. Understandably, during the training process of the business identification model, historical business data in the training data is also encoded to generate corresponding historical encoding vectors. These historical encoding vectors are then input into the large model or neural network model to be trained to complete the training and obtain the business identification model.

[0033] Continue to refer to Figure 1 Step S3 involves selecting one step generation model from multiple step generation models as the current model based on the business category. It is understood that each step generation model is trained based on historical business data and historical operation plans related to the corresponding business category. Therefore, the business category corresponding to the current model is the same as the business category corresponding to the business data, enabling the current model to effectively process the business data and obtain accurate and reliable output data. Step S4 involves determining multiple business steps corresponding to the business data based on the business data and the current model. In this embodiment, the business data is encoded to form encoded data, which is then input into the current model to obtain the corresponding multiple business steps. In this embodiment, the current model is built and trained based on the Transformer model, and accordingly, the current model includes a self-attention mechanism and multi-head attention. The expression for the self-attention mechanism of the current model is: In the formula For querying the matrix, The key matrix, For value matrices, Key matrix Dimensions This is the normalization function. The expression for the multi-head attention mechanism in the current model is: ,in, In the formula For the first One point of attention, This is a vector concatenation operation. This is to output the weight matrix. In this embodiment, the expression for the current model generating multiple business steps is: In the formula A sequence of steps that includes multiple business steps. The encoded data corresponding to the business data. The length of the step sequence. For the first step in the sequence of steps Each business step For the first All business steps that have been generated prior to this business step.

[0034] Continue to refer to Figure 1 Step S5 involves sorting multiple business steps based on the step dependency data and the sorting algorithm, and determining the sorted business steps as the operation schemes corresponding to the business data. In this embodiment, the step dependency data is used to reflect the constraints imposed on the business steps by nuclear power plant rules and actual nuclear power plant factors. It should be noted that this application does not limit the specific objects of the constraints reflected by the step dependency data. In some embodiments, the step dependency data is used to reflect the constraints imposed on the business steps by nuclear power plant rules, while in other embodiments, the step dependency data is used to reflect the constraints imposed on the business steps by actual nuclear power plant factors. Further reference... Figure 5 Step S5 includes the following sub-steps. Step S51 involves sorting the multiple business steps according to a sorting algorithm to obtain the sorted business steps. In this embodiment, the sorting algorithm includes the LambdaMART sorting algorithm. Accordingly, the expression for the sorting function in the LambdaMART sorting algorithm is: In the formula For the sorted multiple business steps, The number of decision trees in the LambdaMART sorting algorithm. For the first The weights corresponding to each decision tree For the first The prediction results of a decision tree for multiple business steps. The expression for the gradient boosting process in the LambdaMART sorting algorithm is: In the formula For the first The model output after one iteration For the first The model output after one iteration The learning rate is used. The LambdaMART sorting algorithm can be optimized using the NDCG (Normalized Discounted Cumulative Gain) metric. The NDCG metric is calculated as follows: ,in, In the formula For normalized loss indicators, The revenue generated by ranking the current model. For the maximum benefit under ideal conditions, The total number of location samples. For the first The true relevance level of each location sample.

[0035] Continue to refer to Figure 5 Step S52 involves performing a sorting verification on the sorted business steps based on the general dependency graph and the hierarchical dependency graph corresponding to the business level where the business data resides, resulting in a sorted set of business steps. In this embodiment, the hierarchical step type can be determined for each business category within each business level, thus using the step type as a node in the graph and the constraints between two related hierarchical step types as edges between corresponding nodes, thereby constructing the corresponding hierarchical dependency graph. Similarly, in this embodiment, the general step type of each business step in a nuclear power plant can be determined, using each general step type as a node in the graph and the constraints between two related general step types as edges between corresponding nodes, thus constructing the corresponding general dependency graph. In other words, the general step type and hierarchical step type corresponding to each business step can be determined, thereby determining the constraints between pairs of business steps, and then performing a sorting verification on the sorted business steps based on these constraints to obtain a sorted set of business steps. The sorting verification operation in this embodiment includes one or more of the following: adjusting the order of business steps, adding business steps, and deleting business steps. Furthermore, in the construction of the general dependency graph and hierarchical dependency graph in this embodiment, corresponding constraints are set according to the risk level of the business steps to reduce conflicts between related business steps, thereby improving the reliability and effectiveness of the execution of multiple business steps that have been ordered. For example, when determining the multiple business steps corresponding to the isolation operation of a nuclear power plant, the corresponding order of each business step in the isolation operation is determined according to the risk level from high to low.

[0036] Continue to refer to Figure 6In this embodiment, the generation method 100 further includes the following steps. Step S6 is to obtain the validity score corresponding to the user execution plan and the operation plan corresponding to the business data. In this embodiment, the user execution plan is the final plan in which the user performs multiple business steps on the business to be executed corresponding to the business data. In this embodiment, the user can generate the user execution plan by referring to the operation plan. That is, the user can adjust the operation plan and use the adjusted operation plan as the user execution plan. Furthermore, the user can determine whether to modify, delete, or perform other adjustment operations on the business steps according to the risk level corresponding to each business step in the operation plan, so as to further improve the reliability and effectiveness of the user execution plan. In this embodiment, the validity score is used to reflect the effectiveness of the operation plan. That is, the higher the validity score, the more effective the operation plan. In this embodiment, the user can manually score the operation plan to determine the validity score of the operation plan. It should be noted that this application does not limit the method of determining the validity score. In some embodiments, the difference between the user execution plan and the operation plan is calculated to obtain the difference value, and the difference value is used as the validity score. For example, the semantic similarity between the user execution plan and the operation plan is calculated, and the semantic similarity is used as the validity score. Understandably, when using the user execution plan as a benchmark for the effectiveness of the operation plan, the higher the effectiveness score of the operation plan, the more similar the operation plan is to the user execution plan.

[0037] Continue to refer to Figure 6 Step S7 involves determining the correction amount corresponding to the business data based on the user execution plan and the corresponding operation plan. In this embodiment, the expression for the correction amount is: In the formula Execute the plan for the user. As an operational plan, This is a correction amount. In this embodiment, the correction amount is... This refers to the set of business steps where the user execution plan and the operation plan differ. For example, if the operation plan includes steps 1 to 5, and the corresponding user execution plan includes steps 1 to 3 and steps 5 to 6, then the correction amount... This includes steps 4 and 6. Further, the correction amount... It is also used to record the previous and next business steps corresponding to the business steps that differ during execution, thereby preserving the corresponding process logic of the business steps that differ, so as to improve the amount of information in the training samples built later, and thus improve the effectiveness of the training samples.

[0038] Continue to refer to Figure 6Step S8 involves constructing training samples based on business data, corresponding correction amounts, validity scores, and user execution plans. Step S9 involves retraining the corresponding business identification model and / or step generation model based on the training samples. It is understood that if the difference between the operation plan and the user execution plan in the training samples corresponds to an incorrect business category identification, then the training samples can be used to retrain the business identification model to improve its accuracy. Similarly, if the difference between the operation plan and the user execution plan in the training samples corresponds to inaccurate business steps, then the training samples can be used to retrain the corresponding step generation model to improve the generation accuracy of the step generation model corresponding to the business category of the training samples.

[0039] The generation method 100 in this embodiment addresses the complexity of objects involved in different business scenarios within a nuclear power plant by dividing the business scenarios into multiple different business levels, thereby effectively distinguishing multiple objects with the same function but different levels of complexity within the nuclear power plant. Based on this, according to the historical business data and historical operation plans corresponding to each business level, the multiple business scenarios under that business level are classified, thereby determining multiple business categories corresponding to each business level. Thus, each business category can be further trained using the corresponding historical business data and historical operation plans to generate a step generation model corresponding to that business category. Furthermore, for all business categories, combined with the corresponding historical business data, a business identification model can be further trained. Through the above settings, in this embodiment, by executing step S2, the business category corresponding to the business to be executed can be effectively determined using the business identification model and the business data corresponding to the business to be executed. Based on this, by executing steps S3 and S4, the step generation model corresponding to that business category can be selected as the current model, and by using the current model and business data, multiple business steps corresponding to the business to be executed can be further obtained. Because the various business scenarios in nuclear power plants are subdivided according to business hierarchy and business category, the current model can effectively identify business data and generate accurate business steps. In this embodiment, by executing step S5, multiple business steps are initially sorted using a sorting algorithm. Further, the step dependency data associated with the business to be executed is used to verify and adjust the initially sorted business steps, ensuring that the sorted business steps better conform to the relevant rules and actual needs of the nuclear power plant. Furthermore, in this embodiment, by executing steps S6 to S9, training samples are constructed using user execution plans generated by the user based on the operation plan to complete the business to be executed, along with related data. This allows for the adjustment and updating of the business identification model and the corresponding step generation model, thereby continuously improving the accuracy of the step generation model and business identification model with use and enabling continuous adjustment as the real-time needs of the nuclear power plant change.

[0040] An embodiment of this application also proposes a method such as Figure 7 The nuclear power plant operation scheme generation system 200 shown is hereinafter referred to as generation system 200. According to... Figure 7 The generation system 200 may include an internal communication bus 21, a processor 22, a read-only memory (ROM) 23, a random access memory (RAM) 24, and a communication port 25. When applied to a personal computer, the generation system 200 may also include a hard disk 26.

[0041] The internal communication bus 21 enables data communication between components of the generation system 200. The processor 22 can perform judgments and issue prompts. In some embodiments, the processor 22 may consist of one or more processors. The communication port 25 enables data communication between the generation system 200 and external systems. In some embodiments, the generation system 200 can send and receive information and data from a network via the communication port 25.

[0042] The generation system 200 may also include different types of program storage units and data storage units, such as hard disk 26, read-only memory (ROM) 23, and random access memory (RAM) 24, capable of storing various data files used for computer processing and / or communication, as well as possible program instructions executed by processor 22. The processor executes these instructions to implement the main part of the method. The results of processor processing are transmitted to the user equipment via a communication port and displayed on the user interface.

[0043] In addition, this application also proposes a computer-readable medium storing computer program code, which, when executed by a processor, implements the above-described method for generating nuclear power operation schemes.

[0044] The basic concepts have been described above. Obviously, for those skilled in the art, the above disclosure is merely illustrative and does not constitute a limitation of this application. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this application. Such modifications, improvements, and corrections are suggested in this application, and therefore remain within the spirit and scope of the exemplary embodiments of this application.

[0045] Furthermore, this application uses specific terms to describe embodiments of the application. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic related to at least one embodiment of the application. Therefore, it should be emphasized and noted that "an embodiment," "one embodiment," or "an alternative embodiment" mentioned twice or more in different locations in this specification do not necessarily refer to the same embodiment. In addition, certain features, structures, or characteristics in one or more embodiments of the application can be appropriately combined.

[0046] Similarly, it should be noted that, in order to simplify the description of the present application and thus aid in the understanding of one or more embodiments, the foregoing description of the embodiments of the present application sometimes combines multiple features into a single embodiment, drawing, or description thereof. However, this disclosure method does not imply that the subject matter of the present application requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of the single embodiments disclosed above.

[0047] Some aspects of this application can be executed entirely by hardware, entirely by software (including firmware, resident software, microcode, etc.), or by a combination of hardware and software. The aforementioned hardware or software may be referred to as a "data block," "module," "engine," "unit," "component," or "system." The processor may be one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DAPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, or combinations thereof. Furthermore, aspects of this application may manifest as computer products residing in one or more computer-readable media, including computer-readable program code. For example, computer-readable media may include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic tapes, etc.), optical discs (e.g., compressed CDs, digital multifunction DVDs, etc.), smart cards, and flash memory devices (e.g., cards, sticks, key drives, etc.).

[0048] A computer-readable medium may contain a propagated data signal containing computer program code, for example, on baseband or as part of a carrier wave. This propagated signal may take various forms, including electromagnetic, optical, and so on, or suitable combinations thereof. A computer-readable medium can be any computer-readable medium other than a computer-readable storage medium, which can be connected to an instruction execution system, apparatus, or device to enable communication, propagation, or transmission of a program for use. The program code located on the computer-readable medium can be propagated through any suitable medium, including radio, cable, fiber optic cable, radio frequency signals, or similar media, or any combination of the above media.

[0049] Although this application has been described with reference to specific embodiments, those skilled in the art should recognize that the above embodiments are only used to illustrate this application, and various equivalent changes or substitutions can be made without departing from the spirit of this application. Therefore, any changes or modifications to the above embodiments within the essential spirit of this application will fall within the scope of the claims of this application.

Claims

1. A method for generating a nuclear power plant operation scheme, characterized in that, Applied to nuclear power plants, the generation method includes: Acquire business data, step dependency data, a trained business identification model, and multiple trained step generation models, wherein the business data includes the business object, business reason, business scope, and affected area corresponding to the nuclear power plant; Based on the business data and the business identification model, determine the business category corresponding to the business data; Based on the business category, select one of the step generation models from the multiple step generation models as the current model; Based on the business data and the current model, determine multiple business steps corresponding to the business data; Based on the step dependency data and the sorting algorithm, the multiple business steps are sorted, and the sorted multiple business steps are determined as the operation schemes corresponding to the business data. The step dependency data is used to reflect the constraints of nuclear power plant rules and / or actual nuclear power plant factors on the business steps.

2. The method for generating a nuclear power plant operation scheme as described in claim 1, characterized in that, The step dependency data includes a general dependency graph and multiple hierarchical dependency graphs. The steps of acquiring business data, step dependency data, a trained business recognition model, and multiple trained step generation models further include: Determine the multiple business levels corresponding to the nuclear power plant; Based on multiple historical business data and corresponding historical operation schemes of the nuclear power plant, the general dependency graph and the hierarchical dependency graph corresponding to each business level are constructed.

3. The method for generating a nuclear power plant operation scheme as described in claim 2, characterized in that, The steps for acquiring business data, step dependency data, a trained business recognition model, and multiple trained step generation models also include: Based on the multiple business levels, the multiple historical business data, and the corresponding historical operation schemes, multiple business categories are determined; For each of the business categories, an initial generation model is trained based on the historical business data and historical operation schemes corresponding to the business category, to obtain the trained initial generation model; For each business category, the corresponding completed initial generation model is used as the step generation model corresponding to that business category.

4. The method for generating a nuclear power plant operation scheme as described in claim 3, characterized in that, For each of the business categories, the step of training the initial generation model based on the historical business data and historical operation schemes corresponding to the business category to obtain the trained initial generation model further includes: The historical business steps in the historical operation plan are standardized according to the step dictionary to obtain standardized historical business steps, wherein the standardized historical business steps include action type, operation object, parameter constraints and risk level; For each business category, the initial generation model is trained based on the historical business data and standardized historical business steps corresponding to the business category, to obtain the trained initial generation model.

5. The method for generating a nuclear power plant operation scheme as described in claim 2, characterized in that, The step of sorting the multiple business steps according to the step dependency data and the sorting algorithm, and determining the sorted multiple business steps as the operation scheme corresponding to the business data, further includes: The sorting algorithm is used to sort the multiple business steps to obtain the sorted multiple business steps; Based on the general dependency graph and the hierarchical dependency graph corresponding to the business level where the business data is located, the sorted business steps are sorted and verified to obtain the sorted business steps.

6. The method for generating a nuclear power plant operation scheme as described in claim 1, characterized in that, The step of determining the business category corresponding to the business data based on the business data and the business identification model further includes: The business data is encoded to obtain an encoding vector; The encoded vector is input into the business identification model to obtain the business category corresponding to the business data.

7. The method for generating a nuclear power plant operation scheme as described in claim 6, characterized in that, The calculation expression for the business category is: ， ， In the formula The business category corresponding to the business data. Business Category The corresponding rating The length of the label sequence. The label transition matrix, For tags The business identification model outputs probabilities.

8. The method for generating a nuclear power plant operation scheme as described in claim 1, characterized in that, The generation method further includes: Obtain the user execution plan corresponding to the business data and the validity score corresponding to the operation plan; Based on the user execution plan and the operation plan corresponding to the business data, determine the correction amount corresponding to the business data; Training samples are constructed based on the business data and the corresponding correction amount, the validity score, and the user execution plan; The corresponding business identification model and / or the model generated in the step are retrained based on the training samples.

9. A system for generating nuclear power plant operation schemes, characterized in that, include: Memory is used to store instructions that can be executed by the processor; as well as A processor for executing the instructions to implement the method for generating a nuclear power operation scheme as described in any one of claims 1-8.

10. A computer-readable medium storing computer program code, characterized in that, The computer program code, when executed by a processor, implements the method for generating a nuclear power operation scheme as described in any one of claims 1-8.

Images