Reactor core refueling method and device, storage medium and electronic equipment

Abstract

The invention relates to a reactor core refueling method and device, a storage medium and electronic equipment, and relates to the technical field of nuclear, and the method comprises the following steps: firstly, obtaining a reactor core picture book and a refueling program text of a reactor; the refueling program text is subjected to refueling text analysis, a named entity in the refueling program text is obtained, and the named entity comprises a refueling operation object; according to the named entity and a refueling rule base, generating a refueling operation instruction corresponding to the refueling program text; and finally, based on the reactor core picture book, executing a refueling simulation process of the refueling operation object according to the refueling operation instruction, and obtaining a reactor core loading picture after the refueling operation instruction is executed. According to the scheme, on the basis of the reactor core picture book of the reactor, the refueling simulation process of the refueling operation object in the named entity is overlaid according to the refueling operation instruction, manual participation in the refueling simulation process is not needed, the time of single-time refueling simulation operation is shortened, and the refueling simulation efficiency is improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of nuclear technology, and particularly relates to a reactor core refueling method and device, a storage medium and an electronic device. BACKGROUND

[0002] The research reactor operation cycle needs to go through a series of steps such as shutdown cooling, refueling preparation, and start-up, among which the refueling operation is one of the important links to ensure the continuous safe and efficient operation of the research reactor. Usually, a simulation refueling operation is needed before the actual refueling operation to verify the accuracy and feasibility of the refueling operation, thereby reducing the risks and hidden dangers that may occur in the field operation.

[0003] At present, the related technology usually prepares a refueling operation card by a professional technician, specifies the refueling operation of the full load of the research reactor operation cycle, and needs to organize relevant personnel to conduct compliance inspection. However, the manual simulation refueling method in the related technology has the problem of long time consumption of single simulation refueling operation and low simulation refueling efficiency. SUMMARY

[0004] Therefore, the present disclosure provides a reactor core refueling method and device, a storage medium and an electronic device, which mainly aims to solve the technical problem of long time consumption of single simulation refueling operation and low simulation refueling efficiency of the manual simulation refueling method in the related technology.

[0005] According to a first aspect of the present disclosure, a reactor core refueling method is provided, which comprises:

[0006] obtaining a reactor core atlas and a refueling program text of a reactor;

[0007] performing refueling text analysis on the refueling program text to obtain named entities in the refueling program text, the named entities including refueling operation objects;

[0008] generating refueling operation instructions corresponding to the refueling program text according to the named entities and a refueling rule library;

[0009] performing a refueling simulation process of the refueling operation objects according to the refueling operation instructions based on the reactor core atlas, and obtaining a reactor core loading map after the refueling operation instructions are executed.

[0010] According to a second aspect of the present disclosure, a reactor core refueling device is provided, which comprises:

[0011] an obtaining module, configured to obtain a reactor core atlas and a refueling program text of a reactor, and perform refueling text analysis on the refueling program text to obtain named entities in the refueling program text, the named entities including refueling operation objects;

[0012] The generation module is used to generate the material change operation instructions corresponding to the material change program text based on the named entities and the material change rule library;

[0013] The execution module is used to execute the refueling simulation process of the refueling operation object according to the refueling operation instruction based on the core diagram, and to obtain the core loading diagram after the refueling operation instruction is executed.

[0014] According to a third aspect of this disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to said at least one processor; wherein the memory stores instructions executable by said at least one processor, said instructions being executed by said at least one processor to enable said at least one processor to perform the method of the first aspect described above.

[0015] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method of the first aspect described above.

[0016] Compared with related technologies, the reactor core refueling method, apparatus, storage medium, and electronic equipment disclosed herein first obtain the reactor core atlas and refueling procedure text; perform refueling text parsing on the refueling procedure text to obtain named entities in the refueling procedure text, the named entities including refueling operation objects; then generate refueling operation instructions corresponding to the refueling procedure text based on the named entities and the refueling rule base; finally, based on the reactor core atlas, execute the refueling simulation process of the refueling operation object according to the refueling operation instructions, and obtain the reactor core loading diagram after the execution of the refueling operation instructions. The scheme disclosed herein can parse the refueling procedure text of different reactors, obtain the named entities in the text, and combine it with the refueling rule base to convert the refueling procedure text written in natural language into executable refueling operation instructions. Then, based on the reactor core diagram, the refueling simulation process is superimposed on the refueling operation objects in the named entities according to the refueling operation instructions. Finally, the core loading diagram after the refueling operation instructions are executed is obtained. No manual intervention is required to simulate the refueling process, reducing the time of a single simulated refueling operation and improving the efficiency of simulated refueling. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0018] To more clearly illustrate the technical solutions in the embodiments or related technologies of this disclosure, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of a core refueling method provided in an embodiment of this disclosure;

[0020] Figure 2 A flowchart illustrating an example provided in an embodiment of this application is shown;

[0021] Figure 3 A flowchart illustrating an example provided in an embodiment of this application is shown;

[0022] Figure 4 A flowchart illustrating an example provided in an embodiment of this application is shown;

[0023] Figure 5 A flowchart illustrating an example provided in an embodiment of this application is shown;

[0024] Figure 6 This is a schematic diagram of a reactor core refueling device provided in an embodiment of the present disclosure. Detailed Implementation

[0025] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description. It should be noted that, unless otherwise specified, the embodiments of this disclosure and the features described therein can be combined with each other.

[0026] The following description, with reference to the accompanying drawings, outlines embodiments of a reactor core refueling method, apparatus, storage medium, and electronic equipment.

[0027] In some embodiments, the reactor operation cycle involves a series of steps, including reactor shutdown and cooling, refueling preparation, and reactor startup. During the shutdown period, preparatory work is required, such as the design and verification of the core loading scheme. The core loading scheme design involves detailed considerations regarding fuel element burnup and relocation, the number and arrangement of target components, the number and location of beryllium blocks, aluminum blocks, and stainless steel blocks, and the position of control rods. The core loading scheme provides the basis for the design of the refueling operation card and core loading diagram. Currently, the refueling operation card and loading diagram are prepared by professional technicians, and relevant personnel must conduct conformity checks to confirm the correctness and operability of the refueling procedure.

[0028] The refueling operation card specifies the refueling operations for a full load of the research reactor during operation. It mainly includes refueling preparation, sample removal from the reactor, unit removal from the reactor, the main refueling procedure, sample loading, and criticality push-in. Operators must strictly follow the steps on the operation card. Typically, a simulated refueling operation is performed before the actual refueling operation to verify the accuracy and feasibility of the operation card, thereby reducing potential risks and hazards during on-site operation.

[0029] Currently, simulated refueling operations are primarily conducted on simulation boards, while manual paper verification is performed simultaneously. This verification process requires checking the loading diagram, loading plan, and operational steps, all three working in tandem, with errors corrected promptly during the process. This method is labor-intensive, cannot record operational trajectories in a timely manner, and makes it difficult to trace back the verification process. Furthermore, the manual step-by-step verification method is prone to parameter misinterpretation due to visual fatigue, posing a risk of misjudgment. It also lacks spatial visualization support, making it difficult to intuitively identify operational conflicts or violations, resulting in low efficiency. For multi-reactor nuclear facilities, this further increases the difficulty of simulated refueling operations. The current digital refueling process still faces the following core bottlenecks:

[0030] First, collaboration among multi-source heterogeneous data is difficult. There is a lack of intelligent correlation mechanisms between basic core information, pool data, core design data, operation card text instructions, and 2D loading diagrams. Reliance on manual conversion easily leads to information distortion and operational deviations. Second, simulated refueling operations lack fault tolerance. The ability to reverse-engineer operational steps has not been established. Correcting errors requires a complete resimulation of the entire process, which is inefficient and lacks the ability to backtrack to specific areas. Third, adapting to multiple reactor types is costly. Component specifications, channel conditions, and coordinate parameters for different reactor types require repeated manual configuration and verification, lacking a unified adaptation framework and program. Finally, there is a break in the transmission of expert experience. Refueling rules rely on manual experience, and a reusable knowledge base has not been formed. Training new personnel is lengthy, and operational consistency is difficult to guarantee.

[0031] This disclosure provides a core refueling method, apparatus, storage medium, and electronic equipment, with the main purpose of solving the technical problems of long time consumption and low efficiency of manual simulated refueling in the current related technologies.

[0032] like Figure 1 As shown, embodiments of this disclosure provide a core refueling method, including:

[0033] Step 101: Obtain the reactor core diagram and refueling procedure text.

[0034] In some embodiments, reactor core diagrams can be drawn based on the basic data of different reactors. The refueling procedure text can be natural language instructions compiled according to a preset format or structure based on different core loading schemes, used to represent the refueling steps corresponding to the core loading scheme. When the core loading scheme is updated, resulting in changes to the refueling logic, this file can be updated and uploaded, thus triggering refueling operation reload and version iteration, eliminating the need for manual re-execution of the refueling steps and improving the efficiency of refueling simulation. The reactors may include, but are not limited to, research reactors, power reactors, production reactors, and propulsion reactors.

[0035] Step 102: Parse the material change program text to obtain the named entities in the material change program text. The named entities include the material change operation object.

[0036] Named entities can be refueling information extracted from the refueling procedure text, including but not limited to refueling operation object, refueling operation type, target object location (target location), etc.; refueling operation object (source object) can be the fuel assembly to be operated, such as component ID, etc.; refueling operation type can be the processing type of the refueling operation object, such as insertion, removal, movement, etc.; target object location can be the target location to which the refueling operation core should be moved according to the refueling operation type, such as transport container A, transport container B, etc.

[0037] In some embodiments, natural language processing techniques can be used to parse the material changing procedure text, such as matching entities in the text based on regular expressions or writing specific rules, or training a model using labeled data to perform named entity recognition, and fine-tuning the model to adapt to the named entity recognition task corresponding to the material changing operation, so as to parse the material changing operation information in the text.

[0038] Step 103: Generate the material change operation instructions corresponding to the material change program text based on the named entities and the material change rule library.

[0039] In some embodiments, the material change rule base may include multiple preset material change rules, such as operation logic rules, safety constraint rules, and sequence constraint rules. Based on the mapping relationship between entities and the material change rule base, the material change rules corresponding to the named entities parsed from the material change operation text can be matched, and corresponding material change operation instructions can be generated according to the rules. The material change operation instructions are executable code instructions. Correspondingly, multiple named entities in the material change operation text can be parsed, and the association relationships between each named entity can be confirmed, and converted into a sequence of material change operation instructions for automatic simulation of the material change process.

[0040] For example, a preset instruction template can be used to select different instruction templates according to the material change operation type, fill in the corresponding positions of the material change operation object and the target object position corresponding to the material change operation type, establish the processing relationship between the material change operation object and the target object position, generate the structured material change operation instruction corresponding to the material change operation type, and simulate the material change operation process.

[0041] Step 104: Based on the core diagram, execute the refueling simulation process of the refueling operation object according to the refueling operation instruction, and obtain the core loading diagram after the refueling operation instruction is completed.

[0042] For example, Canvas technology can be used to draw a reactor core diagram based on the reactor's basic data, such as a static baseline view, to show the initial state of the core. Then, based on the static baseline view, the refueling operation steps corresponding to the refueling operation command can be executed. Specifically, the position of the refueling operation object to be processed in the diagram can be determined, and the object can be moved to the target object position in the diagram according to the refueling operation type. The refueling simulation process of the refueling operation object is completed and displayed. Finally, the core loading diagram after refueling is completed is obtained, thereby obtaining the refueling simulation animation of the refueling operation command overlaid on the core layer. The refueling command is transformed into an intuitive graphical operation process, and the core loading diagram is dynamically updated. This realizes the two-dimensional graphic drawing and animation demonstration of the reactor refueling simulation process, eliminating the need for manual refueling simulation and improving the efficiency of refueling simulation operation.

[0043] Compared with related technologies, this embodiment first obtains the reactor core diagram and refueling procedure text; it then parses the refueling procedure text to obtain named entities, which include refueling operation objects; next, based on the named entities and the refueling rule base, it generates refueling operation instructions corresponding to the refueling procedure text; finally, based on the core diagram, it executes the refueling simulation process of the refueling operation objects according to the refueling operation instructions, and obtains the core loading diagram after the execution of the refueling operation instructions. Through this disclosed solution, the refueling procedure text of different reactors can be parsed to obtain the named entities in the text. Combined with the refueling rule base, the refueling procedure text written in natural language can be converted into executable refueling operation instructions. Then, based on the reactor core diagram, the refueling simulation process of the refueling operation objects in the named entities is superimposed according to the refueling operation instructions, and finally, the core loading diagram after the execution of the refueling operation instructions is obtained. This eliminates the need for manual intervention in the simulated refueling process, reduces the time of a single simulated refueling operation, and improves the efficiency of simulated refueling.

[0044] Based on the technical implementation shown in the above embodiments, in order to further illustrate the specific implementation process of the method in this embodiment, step 101 may optionally include: drawing the static core structure in the core atlas according to the reactor core basic data; receiving a pre-compiled refueling procedure text, the refueling procedure text being adjusted according to the reactor core loading scheme.

[0045] In some embodiments, before drawing the static core structure in the core atlas based on the reactor core basic data, a basic database can be constructed. This basic database can be used for drawing core loading diagrams, pool diagrams, and simulating refueling operations. The basic database may include, but is not limited to, structured stored data such as basic core data, core coordinate data, and basic pool information. The basic core data may include, but is not limited to, facility name, balancing zone, key measurement point code, layout type, shape, size, and color. The core coordinate data may include, but is not limited to, facility name, balancing zone, key measurement point code, core location, and storage location coordinates. The basic pool information may include, but is not limited to, facility name, balancing zone, key measurement point code, storage rack name, storage rack location, and storage rack coordinates. In this way, a reactor basic database can be established, and a parametric template library can be used to pre-set component specifications and core parameters for various reactor types, providing strong data support and flexibility for the system.

[0046] For example, based on the constructed base database, the Canvas technology can be used to draw two-dimensional graphics and animate the core components of the research heap. By creating a Canvas canvas on a web page, the core diagram can be drawn using the Canvas context obtained through JavaScript, based on the basic data of the research heap core, such as component type, coordinates, size, color, etc., and the core diagram can be drawn using functions (methods) such as fillRect, arc, beginPath, stroke, etc., to obtain the static core structure of the research heap.

[0047] Accordingly, based on the current reactor core loading scheme, a refueling procedure text corresponding to the core loading scheme can be prepared in advance. The refueling procedure text can be written according to a preset format, such as "Component A → Transport Container A", which is used to input natural language instructions corresponding to the refueling operation steps.

[0048] Optionally, step 102 may specifically include: using a preset entity recognition model to identify named entities in the material change procedure text according to the grammatical structure of the material change procedure text. The preset entity recognition model is used to identify entities in the material change procedure text according to the overall recognition rules and label transfer constraint rules.

[0049] In some embodiments, the preset entity recognition model can be a BERT-BiLSTM-CRF hybrid model. This model is used to analyze the syntactic structure of the material changeover procedure text, extract named entities, and then map the parsed named entities to predefined operation instructions. For example, the result of parsing the material changeover procedure text "Component A → Transport Bucket A" is: {Operation Type: Move; Operation Object: Component A; Target Location: Transport Bucket A}.

[0050] Specifically, the named entity recognition process requires text data to be labeled before training. For entity recognition labeling tasks, the corpus labeling results include entity category, entity boundary, and length, etc. The BIOES sequence labeling method can be used to standardize the labels within entities. Here, B represents the start of an entity, I represents the inside of an entity, O represents a non-entity, E represents the end of an entity, and S represents a single text character as an entity.

[0051] For example, such as Figure 2As shown, the BERT-BiLSTM-CRF hybrid model consists of three parts: a BERT layer, a BiLSTM layer, and a CRF layer. The BERT layer transforms the original input text sequence of material exchange operations into a numerical vector matrix, obtaining the corresponding low-dimensional embedding vector representation and generating context-sensitive word vectors (e.g., distinguishing "bucket" in "transport bucket A" as a noun rather than a verb). The BiLSTM layer, based on the output of the BERT layer, further learns the temporal features and contextual information of the text sequence, capturing sequence dependencies (e.g., "instruction element A" needs to be identified as a whole rather than disassembled). The CRF layer calculates the concatenated output of the above two modules, decodes it to obtain the final label sequence, and can also perform constrained label transfer, such as ensuring that "B-OBJ" must be followed by "I-OBJ" to avoid illegal sequences. The specific recognition process is as follows:

[0052] In this model, BERT serves as a word embedding layer, an unsupervised, pre-trained language representation model, and can be a bidirectional encoder based on the Transformer model. BERT can convert text characters into one-dimensional vectors, while also capturing the dependencies between text sequences in the material exchange operation text and fusing semantic information from the entire text. The BERT model can obtain context-related information for each character in the material exchange operation text, thus enabling the representation of different meanings of the same character in different sentences or different positions within a sentence. For example, inputting the text "Component A → Transport Bucket A" into a bidirectional Transformer encoder (referred to as "Trm") can output a dynamic character vector W.

[0053] BiLSTM is a recurrent neural network used to process and predict long-term dependencies in sequences. It can be composed of forward LSTM and backward LSTM, capturing bidirectional semantic information corresponding to material exchange operation text to uncover richer semantic features. Unidirectional LSTM models, when processing text, cannot simultaneously consider the semantic relationships between instruction information that is in the reverse order of the text. Therefore, a BiLSTM model is used to merge the forward and backward representations of the input dynamic word vector W to simultaneously capture the information dependencies in both the forward and backward directions of the text, thus providing a more comprehensive understanding of the semantic content of the material exchange operation text.

[0054] After passing through the BiLSTM layer, the output is the score of each character corresponding to each label. Conditional Random Fields (CRF) consider the dependencies and constraints between different labels, and can correct the output of the BiLSTM to ensure the accuracy and validity of the labels. The main constraints of the CRF layer on the label output are: ① Entity labels must begin with "B-"; ② Consecutive entity labels "B-label, I-label, ..., E-label", where labels should be entities of the same class; ③ E represents the end of the entity. CRF learns the transition probabilities between labels and the feature representations of sequence elements, and decodes the fused feature information generated by the aforementioned model to optimize and obtain the correct path that maximizes the joint probability of the entire sequence labels. For example, after the text "Component A → Transport Barrel A" is transformed and labeled by the model, the sequence "B-B1, I-B1, E-B1, O, B-B2, I-B2, I-B2, E-B2" can be generated. Based on this sequence, the named entities in the material changing procedure text can be parsed.

[0055] In some embodiments, Named Entity Recognition (NER) technology is used to identify key entities (such as component ID, target location, etc.) in the material changeover procedure text, analyze the syntactic structure of the material changeover procedure text, extract the material changeover operation type (e.g., "→" represents movement), source object, and target location, and map the parsed entities and operations to predefined operation steps. For example, the entity result after parsing the material changeover procedure text "Component A→Transport Bucket A" is: {Operation Type: Movement; Operation Object: Component A; Target Location: Transport Bucket A}.

[0056] Optionally, before step 103, the method of this embodiment may further include: determining the basic rule set corresponding to the reactor based on the reactor's equipment characteristics and refueling steps; using the test cases corresponding to the basic rule set to verify the rule validity of the basic rule set and obtain the rule coverage of the basic rule set; updating the basic rule set based on the rule coverage of the basic rule set, and constructing a refueling rule library based on the updated basic rule set.

[0057] In some embodiments, a rule base covering spatial transformation, object matching, and safety constraints for all refueling operations can be established based on the reactor's equipment characteristics and all refueling situations. For example, equipment characteristics may include a video inspection platform having only one storage location, a target barrel containing multiple long inner tubes being able to hold multiple long inner tubes, and a transfer container being able to hold multiple components, etc.

[0058] Specifically, spatial rules can be formulated based on the characteristics of the research stack equipment (such as the size of the transfer rack and the capacity of the water tank), and more than 1,200 refueling steps can be sorted out and summarized into 9 basic operation sets (covering movement, position conversion, and safety detection). Examples of basic operation sets are shown in Table 1, which may include: → transport bucket A / bucket transport B, → video inspection platform, → storage water tank, → inspection pool C01, → stacking, → transfer rack, → on the transfer rack, device → on the transfer rack channel, device → channel, and other refueling procedure rules.

[0059] Table 1. Examples of Basic Operation Sets

[0060] Serial number Rule 1 → Transporting bucket A / bucket transporting bucket B 2 → Video inspection platform 3 → Holding tank 4 → Inspection tank C01 5 → In-pile 6 → Conversion stand 7 → On conversion stand 8 Device → On conversion stand channel 9 Device → Channel

[0061] Optionally, conflict resolution rules can be preset, such as temporarily placing the component on the conversion rack if the target location is already occupied.

[0062] Correspondingly, the validity of the basic rule set can be verified. Test cases can be designed for the nine types of basic rule sets, such as a full-link simulation of "transport bucket A → transfer rack → stacking". The rule coverage of the basic rule set is tested by replaying historical material change data. If the rule coverage of the basic rule set is greater than or equal to a preset coverage threshold (e.g., 100%), the basic rule set can be used as the material change rule library. If the rule coverage of the basic rule set is less than the preset coverage threshold, the basic rule set can be reviewed and iterated until the rule coverage is greater than or equal to the preset coverage threshold. The updated and iterated basic rule set is then used as the material change rule library. The preset coverage threshold can be a preset threshold to ensure that all material change operation steps can be mapped to, thereby verifying the validity of the rule library. For example, the material change rule library may include more than 1200 specific rules, which can be used as input to a preset instruction recognition model.

[0063] Optionally, step 103 may specifically include: using a preset instruction recognition model to map named entities to preset operation steps, generating material change operation instructions based on the preset operation steps, wherein the preset instruction recognition model is used to construct a dynamic mapping table between named entities and material change operation steps according to the material change rule library.

[0064] In some embodiments, the preset instruction recognition model may include a material change procedure rule intelligent recognition model. The preset instruction recognition model can construct a dynamic mapping table between component feature vectors (type / coordinates / size / color) and operation steps based on the material change rule library, so as to realize the automatic conversion between natural language instructions (such as "component A → transport bucket A") and machine code.

[0065] Optionally, the preset instruction recognition model may include a preset entity recognition model to identify named entities in the material changeover program text during the material changeover instruction conversion process.

[0066] For example, such as Figure 3The diagram illustrates the intelligent recognition process for material changeover procedures. Users can input natural language commands (such as "Component A -> Location B") through the material changeover procedure file. The entity recognition model parses the material changeover procedure file to obtain the structured named entities corresponding to the text. For example, the BERT-BiLSTM-CRF model is used to identify the operation type (such as "move"), operation object (such as "Component A"), and target location (such as "Location B") in the command. Through a logic mapper, the parsed entities and operation types are mapped to predefined operation steps to generate machine-executable code as the material changeover operation command.

[0067] In this way, an intelligent recognition model for material changeover procedures can be built based on multiple sets of basic rules and multiple specific rules. This model can intelligently parse the material changeover procedure text and automatically convert the complex rules originally described in natural language into executable code instructions, thus realizing the automation and intelligence of material changeover simulation.

[0068] Optionally, step 104 may specifically include: in response to receiving a refueling operation instruction, determining the graphic object of the refueling operation in the core atlas; based on the refueling operation type and the target object position, updating the graphic coordinates of the graphic object in the corresponding canvas of the core atlas frame by frame through a timer, generating a movement animation of the image object, the movement animation being used to display the refueling simulation process of the refueling operation object.

[0069] In some embodiments, the reactor core atlas can be drawn using Canvas technology. During initialization, reactor core design data (including component type, coordinates, size, and color) is loaded, and the static reactor core structure is drawn in one go using Canvas, avoiding repeated rendering during dynamic demonstrations. In response to refueling operation commands received from the refueling procedure file, the graphical objects of the fuel elements to be moved in the refueling operation commands are located on the Canvas canvas. A JavaScript timer updates the coordinates of these elements on the canvas frame by frame, creating a smooth movement animation to display the refueling simulation process of the refueling operation objects, thus realizing the dynamic changes in the reactor core atlas. This utilizes Canvas technology to construct an intuitive dynamic visualization interface, enabling two-dimensional graphical drawing of reactor core components and animated demonstration of the refueling process. This interface not only clearly displays the movement trajectory of components but also dynamically identifies rule conflict points in real time, allowing operators to intuitively grasp the status of the refueling process.

[0070] Optionally, the method in this embodiment may further include: obtaining core layout data corresponding to the core loading diagram after the refueling operation object has been moved; comparing the core layout data with the reactor's preset layout data to obtain a verification result of the core layout data, wherein the preset layout data is determined according to the reactor's core loading scheme; if there is a deviation in the verification attribute in the verification result, updating the refueling rule base and / or the preset entity recognition model according to the deviation data of the verification attribute.

[0071] In some embodiments, after the simulated refueling is completed according to the parsed operation instructions, the core layout data corresponding to the core loading diagram can be automatically extracted. The core layout data may include the type and coordinates of all components. Then, the core layout data is compared with the preset layout data corresponding to the core loading scheme. The consistency between the composite core loading design scheme and the refueling operation can be verified. Specifically, it can be verified whether the key attributes (such as coordinates and type) in the layout data are completely matched, output a consistency report, and mark the components with deviations and specific differences.

[0072] Correspondingly, based on the verification results, the operational steps that deviated can be viewed, along with the corresponding deviation data. For example, for operational steps that frequently deviate (such as coordinate offsets of specific movement commands), the associated material changing rule entries can be automatically analyzed. Based on historical deviation data, the parameters in the rule base (such as position tolerance and operation priority weights) can be adjusted. Furthermore, the erroneous command samples in the deviation report (such as command texts that fail to recognize entities) can be used to retrain the BERT-BiLSTM-CRF model. For strengthening the feature weights of easily confused entities (such as "transport bucket A" and "transport bucket B"), the label transfer constraint rules of the CRF layer can be optimized to reduce the occurrence of illegal sequences.

[0073] Optionally, the method in this embodiment may further include: performing conflict interception and dynamic rule verification in each step of the simulated material change operation to detect potential errors in a timely manner.

[0074] Through a verification and feedback optimization layer, the system continuously collects operational data and optimizes the parameters of the refueling rule base and intelligent semantic parsing algorithm. After the simulated refueling is completed, it automatically verifies whether the final core loading scheme meets the preset design requirements, effectively ensuring the accuracy of the simulation results. Simultaneously, the system supports local corrections to the simulation results, avoiding the drawbacks of traditional methods that require repeating the entire simulation process due to a single point of error. This significantly saves rework time and continuously improves the system's adaptability (compatible with different reactor types / rules) and the accuracy of rule recognition and execution. By integrating a highly efficient automatic rule verification module, the system completely replaces the traditional manual verification process, drastically reducing the time for a single simulated refueling operation from approximately 2 hours to less than 5 minutes, improving efficiency by more than 20 times and greatly shortening the simulated refueling operation time.

[0075] For example, such as Figure 4 As shown, a method for optimizing intelligent simulation refueling in a research reactor is illustrated, which may include the following steps:

[0076] S1 Data Preparation Phase: The system can integrate multiple types of core data, such as core basic data (including nuclear facility physical parameters and spatial coordinates), core design data (including loading schemes and physical characteristic parameters), pool information (storage rack configuration and capacity), loading schemes (assembly layout design), and simulated refueling schemes (operation command sequences), etc. The specific steps are as follows:

[0077] Data structuring: Based on the research of the reactor core physical structure, the parameters (size / shape / position) of the reactor core components are converted into computable electronic data, and key information such as the location and capacity of the storage racks are abstracted according to the actual layout of the water tank;

[0078] Design data association: Based on the core input parameters, the loading scheme is prepared. The system automatically triggers the association process to bind the loading scheme with the core database. The loading scheme data is automatically identified by the system and persistently stored in the design database.

[0079] Upload a pre-compiled material change procedure file. The system will automatically parse the operation steps and store the instruction sequence. When the material change logic changes, the file update will trigger data reload and version iteration.

[0080] This stage ultimately outputs standardized, traceable core data and a refueling procedure library, providing a complete data base for simulating refueling operations. All data changes are controlled through a database to ensure consistency, guaranteeing that the input source for subsequent simulation verification is accurate and reliable.

[0081] S2 constructs the basic database: Based on the basic data output by S1 (including structured storage of core basic data management, core coordinate data management, and basic information of the water tank), structured storage is performed. The specific steps are as follows:

[0082] Data standardization processing: Logical relationships are established among three types of data through facility name, balance zone, and key measurement point code. Coordinate data is converted into a standard coordinate system through a unified unit of measurement (coordinate: millimeter; size: millimeter × millimeter) and a standardized coding system. For example, {facility A, balance zone A, key measurement point code A, B01, 100, 200} means that the coordinates of the core position B01 are (X: 100mm, Y: 200mm).

[0083] Data integrity verification: The integrity of basic data fields is checked through database design and system input verification control; the logical rationality of coordinates is verified according to the core design boundary to ensure that the coordinates are within the core design boundary; and the association and matching of basic data and coordinate data are achieved through cross-table consistency verification.

[0084] Data storage management: Establish different database tables, ensure the relationship between data through foreign keys, record data status and version by setting timestamp and data tag fields, and ensure data security through access control;

[0085] This stage forms a structured database, and multiple verification mechanisms ensure the integrity of the data, providing accurate and reliable data input for subsequent core catalog production.

[0086] S3 constructs a rematerial changeover rule base: Referring to the fundamental principles that each target barrel, conversion rack, transport barrel A, transport barrel B, and video inspection platform can only have one storage location; and that when long inner tubes are inserted into the target barrel, one target barrel can hold multiple long inner tubes, and a transfer container can hold multiple components, nine basic rule sets are defined, including: → Transport barrel A / barrel transport barrel B, → Video inspection platform, → Storage pool, → Inspection pool C01, → Stacking, → Conversion rack, → On the conversion rack, Device → Channel on the conversion rack, Device → Channel, etc., basically covering all rematerial changeover scenarios in the research stack. A rule base covering spatial transformation, object matching, and safety constraints for all rematerial changeover operations is established.

[0087] S4 Material Changeover Procedure Rule Intelligent Recognition Model: Based on the material changeover rule library, a dynamic mapping table is constructed between component feature vectors (type / coordinates / size / color) and operation steps to achieve automatic conversion between natural language commands (such as "component A → transport container A") and machine code. The specific steps are as follows:

[0088] Enter the material change operation command, such as "Component A → Transport Container A";

[0089] Named entity recognition is performed using a BERT-BiLSTM-CRF hybrid model. By analyzing the syntactic structure, the operation type (e.g., "→" represents movement), the source object (component A), and the target location (bucket A), a structured entity is generated for the instruction "component A → bucket A": {Operation type: movement, Operation object: component A, Target location: bucket A.}.

[0090] The parsed structured entities and operation types are logically mapped to predefined operation steps, and executable machine code is output. For example, the ID of component A corresponds to the coordinates (X component A, Y component A), and "→" corresponds to the operation movement (MOV). The ID of transport bucket A corresponds to the coordinates (X transport bucket A, Y transport bucket A). The generated code instruction is: MOV((X component A, Y component A), (X transport bucket A, Y transport bucket A)).

[0091] In this way, the material changing operation instructions described in natural language (such as "component A → transport bucket A") are converted into standardized code instructions that can be executed by the machine, such as MOV((X component A, Y component A), (X transport bucket A, Y transport bucket A)), providing a flow of operation instructions that can be directly called for subsequent simulated material changing dynamic demonstrations.

[0092] Dynamic generation of S5 reactor core diagrams: Based on the basic database built by S2, Canvas technology is used to realize the two-dimensional graphic drawing and animation demonstration of the reactor core components. The specific steps are as follows:

[0093] Initial graphic rendering: Match geometric templates (rectangle, circle, polygon) according to component type, accurately lay out in the canvas space according to coordinate parameters, and draw the core graphic using JavaScript data from the Canvas context. Use functions such as fillRect, arc, beginPath, and stroke to draw the core graphic. Among them, the fillRect function can be used to draw a filled rectangle, the arc function can be used to add an arc path, the beginPath function can be used to create a new path, and the stroke function can be used to draw the border of the current path according to the current line style. These functions can be used together to complete complex drawing operations. Component categories (such as components, holes, target barrels, conversion racks, etc.) can also be distinguished by color.

[0094] Dynamic Graphic Updates: A timer is used to acquire data and update the graphics on the Canvas, enabling dynamic changes to the core diagram. A data monitoring channel is established to acquire component state changes in real time. A time increment algorithm is used to smoothly transition the component movement path, triggering a partial redraw when the component state changes.

[0095] The core diagram generated in this stage serves as a static reference view, providing a visual operational platform for subsequent simulated refueling dynamic demonstrations. This diagram is drawn based on precise parameters (coordinates, dimensions, component types) from the basic database, fully presenting the initial state of the core and becoming a spatial reference system for component movement and state changes during dynamic demonstrations.

[0096] S6 Simulated Refueling Dynamic Demonstration: Combining the operation command flow of the simulated refueling operation generated by S4 with the core atlas baseline view generated by S5, the simulated refueling operation can be superimposed on the execution effect of the operation command on the actual core layout, achieving a seamless connection from static design to dynamic verification. The specific steps are as follows:

[0097] Static structure loading: The system loads the core atlas generated by S5, draws a complete static core loading diagram at once through the drawing engine, and establishes a mapping table between component coordinates and graphic objects;

[0098] Dynamic demonstration phase: Based on the operation command stream of the simulated refueling operation generated by S4, locate the corresponding graphic object of the core element to be moved in the core diagram. Update the element coordinates frame by frame using a timer to create a smooth movement animation;

[0099] After the simulation of material change dynamics is completed in this stage, the final state of all components in the drawing board is automatically captured and used as input for subsequent verification and feedback optimization.

[0100] S7 verification and feedback optimization layer: such as Figure 5 As shown, after the simulated refueling dynamic demonstration is completed, the final state of all components in the drawing board is automatically captured. Deviation analysis is performed on the core loading data after executing the refueling operation command corresponding to the refueling operation text, and on the pre-stored core loading scheme data. The consistency between the composite design scheme and the refueling operation steps is verified. For cases where deviations exist, the adaptability and accuracy of the system can be improved by optimizing the refueling rule base or semantic parsing model parameters. Optimization steps may include the following:

[0101] Rule base optimization: For operation steps that frequently deviate (such as coordinate offset of specific movement commands), automatically analyze the associated material change rule entries, automatically update the rule parameters based on the deviation data, and adjust the rule parameters and operation weights;

[0102] Semantic parsing model optimization: Utilizing erroneous instruction samples from deviation reports (such as instruction texts where entity recognition failed, e.g., splitting "transport bucket A" into a verb + noun in "transport bucket A→B"; illegal sequences of consecutive operators in the instruction "component A→→ conversion rack"), the intelligent recognition model for the S4 material changing procedure rules is retrained. This includes strengthening the contextual feature weights of easily confused entities (such as "transport bucket A" and "transport bucket B"), adding technical terminology protection rules (forcing the recognition of "→" as an operator), and optimizing the label transfer constraint rules of the CRF layer (prohibiting consecutive operator occurrences) to reduce the occurrence of illegal sequences.

[0103] Cyclic testing and verification: After the rule base and semantic parsing model are optimized, the system triggers full-scale testing and verification. This test uses inconsistency examples in historical deviation analysis as the test benchmark. The test executes simulated material change operations covering all deviation types. If the coverage reaches the threshold, the optimized content is updated to the production environment. If it does not reach the threshold, the uncovered cases are located and a new round of optimization is triggered to ensure that the optimized system can fully adapt to existing problem scenarios. After passing the cyclic testing and verification, the rule base and semantic parsing model are updated synchronously to continuously improve the system's adaptability and operation parsing accuracy.

[0104] This approach addresses challenges such as difficulties in data collaboration, insufficient fault tolerance in simulated refueling operations, lack of multi-reactor adaptation frameworks, and difficulties in experience transfer. It proposes an innovative solution that integrates rule engines, semantic parsing, and dynamic simulation technologies. It constructs a three-layer closed-loop verification method of "rule construction - dynamic simulation - verification and optimization," which realizes automation and intelligence in simulated refueling of the research reactor. This significantly reduces the uncertainty of on-site refueling operations and provides systematic support for backtracking the historical refueling process and design optimization and improvement of the research reactor.

[0105] Compared with related technologies, this embodiment can parse the refueling procedure text of different reactors, obtain the named entities in the text, and convert the refueling procedure text written in natural language into executable refueling operation instructions by combining it with the refueling rule base. Then, based on the reactor core diagram, the refueling simulation process is superimposed on the refueling operation objects in the named entities according to the refueling operation instructions. Finally, the core loading diagram after the execution of the refueling operation instructions is obtained. No manual intervention is required to simulate the refueling process, reducing the time of a single simulated refueling operation and improving the efficiency of simulated refueling. Furthermore, by using a preset instruction recognition model, named entities are mapped to preset operation steps, and refueling operation instructions are generated based on the preset operation steps. This realizes the conversion of natural language instructions in the refueling procedure text into executable refueling operation code instructions, providing a directly callable operation instruction stream for subsequent simulated refueling dynamic demonstrations. It can also verify the core layout data corresponding to the core loading diagram based on preset layout data, continuously improving the system adaptability and operation parsing accuracy.

[0106] Based on the above Figure 1 The specific implementation of the method shown in this embodiment provides a core refueling device, such as... Figure 6 As shown, the device includes: an acquisition module 31, a generation module 32, and an execution module 33;

[0107] The acquisition module 31 is used to acquire the reactor core diagram and refueling procedure text; to parse the refueling procedure text and acquire the named entities in the refueling procedure text, including the refueling operation object;

[0108] The generation module 32 is used to generate material change operation instructions corresponding to the material change program text based on the named entities and the material change rule library;

[0109] The execution module 33 is used to perform a refueling simulation process of the refueling operation object according to the refueling operation instruction based on the core diagram, and to obtain the core loading diagram after the refueling operation instruction is executed.

[0110] In some examples of this embodiment, the generation module 32 is specifically configured to use a preset instruction recognition model to map named entities to preset operation steps, and generate material change operation instructions based on the preset operation steps. The preset instruction recognition model is used to construct a dynamic mapping table between named entities and material change operation steps according to the material change rule library.

[0111] In some examples of this embodiment, the acquisition module 31 is further configured to determine the basic rule set corresponding to the reactor based on the reactor's equipment characteristics and refueling steps; to perform rule validity verification on the basic rule set using test cases corresponding to the basic rule set, and to obtain the rule coverage of the basic rule set; to update the basic rule set based on the rule coverage of the basic rule set, and to construct a refueling rule library based on the updated basic rule set.

[0112] In some examples of this embodiment, the acquisition module 31 is specifically configured to use a preset entity recognition model to identify named entities in the material change program text according to the grammatical structure of the material change program text. The named entities also include the material change operation type corresponding to the material change operation object and the target object position corresponding to the material change operation object. The preset entity recognition model is used to identify entities in the material change program text according to the overall recognition rules and the label transfer constraint rules.

[0113] In some examples of this embodiment, the execution module 33 is specifically configured to, in response to receiving a refueling operation instruction, determine the graphic object of the refueling operation in the core atlas; based on the refueling operation type and the target object position, update the graphic coordinates of the graphic object in the corresponding canvas of the core atlas frame by frame through a timer, and generate a movement animation of the image object, which is used to display the refueling simulation process of the refueling operation object.

[0114] In some examples of this embodiment, the acquisition module 31 is further configured to acquire the core layout data corresponding to the core loading diagram after the refueling operation object has been moved; compare the core layout data with the preset layout data of the reactor to acquire the verification result of the core layout data, wherein the preset layout data is determined according to the reactor core loading scheme; if there is a deviation in the verification attribute in the verification result, update the refueling rule base and / or the preset entity recognition model according to the deviation data of the verification attribute.

[0115] In some examples of this embodiment, the acquisition module 31 is specifically configured to draw the static core structure in the core atlas based on the reactor core basic data; and to receive a pre-compiled refueling procedure text, which is adjusted according to the reactor core loading scheme.

[0116] Based on the above, Figure 1Accordingly, this embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method. Figure 1 The method shown.

[0117] Based on this understanding, the technical solution of this disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, or portable hard drive) and includes several instructions to cause a computer device (such as a personal computer, server, or network device) to execute the methods of various implementation scenarios of this disclosure.

[0118] Based on the above, Figure 1 The method shown, and Figure 6 To achieve the above objectives, this disclosure also provides an electronic device, comprising a storage medium and a processor; the storage medium for storing a computer program; and the processor for executing the computer program to implement the above-described virtual device embodiments. Figure 1 The method shown.

[0119] Optionally, the aforementioned physical devices may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Wi-Fi interfaces), etc.

[0120] Those skilled in the art will understand that the physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0121] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the aforementioned physical device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software in the information processing physical device.

[0122] Through the above description of the embodiments, those skilled in the art can clearly understand that this disclosure can be implemented by means of software plus necessary general-purpose hardware platforms, or it can be implemented by hardware. Through the solution of this disclosure, this embodiment can parse the refueling procedure text of different reactors, obtain the named entities in the text, and combine it with the refueling rule base to convert the refueling procedure text written in natural language into executable refueling operation instructions. Then, based on the reactor core diagram, the refueling simulation process is superimposed according to the refueling operation instructions on the named entities, and finally, the core loading diagram after the execution of the refueling operation instructions is obtained. No manual intervention is required in simulating the refueling process, reducing the time of a single simulated refueling operation and improving the efficiency of simulated refueling. Furthermore, by using a preset instruction recognition model, named entities are mapped to preset operation steps, and refueling operation instructions are generated based on the preset operation steps. This realizes the conversion of natural language instructions in the refueling procedure text into executable refueling operation code instructions, providing a directly callable operation instruction stream for subsequent simulated refueling dynamic demonstrations. It can also verify the core layout data corresponding to the core loading diagram based on preset layout data, continuously improving system adaptability and operation parsing accuracy.

[0123] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the term "comprising" or any other variations thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0124] The above are merely specific embodiments of this disclosure, enabling those skilled in the art to understand or implement this disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to these embodiments, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for refueling a reactor core, characterized in that, include: Obtain the reactor core diagram and refueling procedure text; The material change procedure text is parsed to obtain named entities in the material change procedure text, and the named entities include material change operation objects. Based on the named entities and the material changing rule library, generate the material changing operation instructions corresponding to the material changing program text; Based on the core diagram, the refueling simulation process of the refueling operation object is executed according to the refueling operation instruction, and the core loading diagram after the refueling operation instruction is completed is obtained.

2. The method according to claim 1, characterized in that, The step of generating the material change operation instruction corresponding to the material change program text based on the named entity and the material change rule library includes: Using a preset instruction recognition model, the named entities are mapped to preset operation steps, and the material change operation instruction is generated based on the preset operation steps. The preset instruction recognition model is used to construct a dynamic mapping table between named entities and material change operation steps according to the material change rule library.

3. The method according to claim 2, characterized in that, Before generating the material change operation instruction corresponding to the material change program text based on the named entity and the material change rule library, the method further includes: Based on the equipment characteristics and refueling procedures of the reactor, determine the basic rule set corresponding to the reactor; Using the test cases corresponding to the basic rule set, the rule validity of the basic rule set is verified, and the rule coverage of the basic rule set is obtained; The basic rule set is updated based on its rule coverage, and the material change rule library is constructed based on the updated basic rule set.

4. The method according to claim 1, characterized in that, The step of parsing the material change procedure text to obtain named entities in the material change procedure text includes: Using a preset entity recognition model, named entities in the material changing program text are identified according to the grammatical structure of the material changing program text. The named entities also include the material changing operation type corresponding to the material changing operation object and the target object location corresponding to the material changing operation object. The preset entity recognition model is used to identify entities in the material change procedure text according to the overall recognition rules and label transfer constraint rules.

5. The method according to claim 4, characterized in that... The refueling simulation process, based on the core diagram and executed according to the refueling operation instructions, for the refueling operation object includes: In response to receiving the refueling operation instruction, determine the graphic object of the refueling operation in the core atlas; Based on the refueling operation type and the target object position, the graphic coordinates of the graphic object in the corresponding drawing board of the core atlas are updated frame by frame by a timer to generate a movement animation of the image object. The movement animation is used to display the refueling simulation process of the refueling operation object.

6. The method according to claim 1, characterized in that, After obtaining the core loading diagram after the refueling operation command has been executed, the method further includes: Obtain the core layout data corresponding to the core loading diagram; By comparing the core layout data with the preset layout data of the reactor, the verification result of the core layout data is obtained. The preset layout data is determined according to the core loading scheme of the reactor. If there is a deviation in the verification attribute in the verification result, the material change rule base and / or the preset entity recognition model are updated according to the deviation of the verification attribute.

7. The method according to claim 6, characterized in that, The acquisition of the reactor core diagram and refueling procedure text includes: Based on the reactor core data, draw the static core structure in the core atlas; Receive a pre-compiled refueling procedure text, which is adjusted according to the reactor core loading scheme.

8. A core refueling device, characterized in that, include: The acquisition module is used to acquire the reactor core diagram and refueling procedure text; The material change procedure text is parsed to obtain named entities in the material change procedure text, and the named entities include material change operation objects. The generation module is used to generate the material change operation instructions corresponding to the material change program text based on the named entities and the material change rule library; The execution module is used to execute the refueling simulation process of the refueling operation object according to the refueling operation instruction based on the core diagram, and to obtain the core loading diagram after the refueling operation instruction is executed.

9. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images