Method and system for quantitatively evaluating decision-making behavior in nuclear emergency situations

Abstract

The application discloses a kind of nuclear emergency under the condition of quantification decision-making behavior evaluation method and system, it is related to emergency management technical field, this method includes: through receiving the decision-making behavior scheme under the condition of nuclear emergency, and carry out prediction, output multiple accompanying emergency events.Construct decision-making behavior prediction tree, the risk assessment of decision-making behavior scheme, output decision-making behavior risk index.Risk assessment is carried out to multiple accompanying emergency events respectively, output multiple accompanying risk index.According to decision-making behavior risk index and multiple accompanying risk index, decision-making behavior prediction tree is quantified and identified, and output the decision-making behavior prediction tree after identification.Through risk fitting to the decision-making behavior prediction tree after identification, generate decision-making behavior evaluation report.The technical problem that nuclear emergency decision-making evaluation method in prior art is difficult to carry out dynamic risk assessment to decision-making behavior, leading to emergency decision-making behavior evaluation accuracy is difficult to improve, influence the reliability of emergency decision-making.

Description

Technical Field

[0001] This application relates to the field of emergency management technology, specifically to a method and system for quantitatively evaluating decision-making behavior in nuclear emergency situations. Background Technology

[0002] In the event of a nuclear emergency, rapid and effective decision-making is crucial for mitigating the impact of the accident and ensuring public safety. In practice, decision-making actions often trigger a series of accompanying emergency events, such as equipment failure, failed evacuation, or radioactive material spread. These accompanying events can significantly increase the difficulty and complexity of emergency response. Therefore, accurately predicting these accompanying events and conducting risk assessments is a key step in optimizing decision-making actions. Traditional risk assessment methods primarily rely on static risk assessments based on historical data, which cannot flexibly respond to real-time data changes, resulting in low prediction accuracy and insufficient reliability of decisions.

[0003] Therefore, existing nuclear emergency decision assessment methods struggle to dynamically assess the risks of decision-making actions, leading to difficulties in improving the accuracy of emergency decision assessments and impacting the reliability of emergency decisions. Summary of the Invention

[0004] This application provides a quantitative decision-making behavior assessment method and system for nuclear emergency situations, solving the technical problem that existing nuclear emergency decision assessment methods struggle to dynamically assess the risks of decision-making behavior, leading to difficulties in improving the accuracy of emergency decision-making behavior assessment and affecting the reliability of emergency decisions. By constructing a decision-making behavior prediction tree and combining it with a feedback adjustment mechanism based on the real-time number of accompanying emergency events, dynamic risk assessment and optimization of accompanying events are achieved, thereby improving the accuracy of decision-making behavior risk assessment in nuclear emergency events and enhancing the reliability of emergency decisions.

[0005] This application provides a method for quantitatively assessing decision-making behavior in a nuclear emergency. The method includes: receiving decision-making behavior plans under a nuclear emergency; predicting the decision-making behavior plans and outputting multiple accompanying emergency events; constructing a decision-making behavior prediction tree by using the multiple accompanying emergency events as child nodes and the decision-making behavior plans under the nuclear emergency as parent nodes; assessing the risk of the decision-making behavior plans and outputting decision-making behavior risk indicators; assessing the risk of each of the multiple accompanying emergency events and outputting multiple accompanying risk indicators; quantitatively labeling the decision-making behavior prediction tree according to the decision-making behavior risk indicators and the multiple accompanying risk indicators, and outputting the labeled decision-making behavior prediction tree; and generating a decision-making behavior assessment report by performing risk fitting on the labeled decision-making behavior prediction tree.

[0006] In the implementation, the decision-making action plan is predicted, and multiple accompanying emergency events are output. The method includes: collecting historical nuclear emergency event samples; constructing an accompanying emergency event prediction model; simulating and predicting the historical nuclear emergency event samples and the accompanying emergency event samples based on the accompanying emergency event prediction model, and obtaining the prediction accuracy; predicting the decision-making action plan based on the accompanying emergency event prediction model, outputting an initial set of accompanying emergency events, and the accompanying probability of each initial accompanying emergency event; and filtering out initial accompanying emergency events with a probability greater than a preset accompanying probability, and outputting these as the multiple accompanying emergency events.

[0007] In the implementation method, an accompanying emergency event prediction model is constructed. This includes: constructing an emergency scenario simulation model; analyzing the evolution process of the historical nuclear emergency event samples using the simulation model; identifying accompanying emergency event samples for each historical nuclear emergency event sample, and accompanying factors for each accompanying emergency event sample, wherein the accompanying factor represents the probability of the accompanying event occurring. The model is then trained based on the historical nuclear emergency event samples, the accompanying emergency event samples, and the accompanying factors for each accompanying emergency event sample. If the prediction accuracy reaches a preset threshold, the accompanying emergency event prediction model is output.

[0008] In the implementation method, the evolution process of the historical nuclear emergency event sample is analyzed using the emergency scenario simulation model, including horizontal evolution analysis and vertical evolution analysis. The horizontal evolution analysis is to simulate the diffusion scenario of the historical nuclear emergency event sample, and the vertical evolution analysis is to simulate the chain scenario of the historical nuclear emergency event sample.

[0009] In the implementation method, a risk assessment is performed on the decision-making behavior scheme, and a decision-making behavior risk index is output. The method includes: performing a partitioned simulation of the decision-making behavior scheme to obtain multiple regions; identifying the personnel protection rate and equipment protection rate in each of the multiple regions, and outputting multiple personnel protection rates and multiple equipment protection rates; and calculating and outputting the decision-making behavior risk index based on the multiple personnel protection rates and multiple equipment protection rates.

[0010] In the implementation method, risk assessments are performed on the multiple accompanying emergency events, and multiple accompanying risk indicators are output. The method includes: collecting samples of accompanying emergency events and samples of accompanying loss indicators based on the event attributes of the multiple accompanying emergency events; establishing an accompanying event risk assessment network based on the accompanying emergency event samples and the accompanying loss indicator samples; and performing a risk assessment on each accompanying emergency event based on the accompanying event risk assessment network, outputting multiple accompanying risk indicators.

[0011] In the implementation, risk fitting is performed on the labeled decision behavior prediction tree. The method includes: dividing the labeled decision behavior prediction tree into hierarchical levels, outputting multi-level child nodes, including first-level child nodes, second-level child nodes, third-level child nodes, ..., Nth-level child nodes; assigning level labels to the multi-level child nodes, outputting multi-level level indicators; using the multi-level level indicators as weights, and the decision behavior risk indicators and the multiple accompanying risk indicators as inputs, performing risk fitting, and outputting a risk fitting index; and generating a decision behavior assessment report based on the risk fitting index.

[0012] In the implementation method, multiple accompanying emergency events are output. The method further includes: obtaining the real-time number of multiple accompanying emergency events; generating an adjustment coefficient based on the ratio of the real-time number to a preset number; and performing feedback adjustment based on the adjustment coefficient when performing risk fitting on the identified decision behavior prediction tree.

[0013] This application also provides a quantitative decision-making behavior assessment system in nuclear emergency situations, including:

[0014] The accompanying event acquisition module is used to receive decision-making action plans in nuclear emergency situations, predict the decision-making action plans, and output multiple accompanying emergency events.

[0015] The behavior prediction tree construction module is used to construct a decision behavior prediction tree by taking the multiple accompanying emergency events as child nodes and the decision behavior scheme under the nuclear emergency as the parent node.

[0016] The decision-making behavior risk assessment module is used to assess the risk of the decision-making behavior plan and output decision-making behavior risk indicators.

[0017] The accompanying emergency risk assessment module is used to assess the risk of the multiple accompanying emergency events and output multiple accompanying risk indicators.

[0018] The identification module is used to quantitatively identify the decision behavior prediction tree according to the decision behavior risk index and the multiple accompanying risk indicators, and output the identified decision behavior prediction tree.

[0019] The assessment report acquisition module is used to generate a decision behavior assessment report by performing risk fitting on the identified decision behavior prediction tree.

[0020] This application proposes a method and system for quantitatively assessing decision-making behavior in nuclear emergency situations. The method receives decision-making behavior plans under nuclear emergency conditions, predicts these plans, and outputs multiple accompanying emergency events. A decision-making behavior prediction tree is constructed, using these accompanying emergency events as child nodes and the decision-making behavior plans as parent nodes. Risk assessments are performed on the decision-making behavior plans, outputting decision-making behavior risk indicators. Risk assessments are also performed on each of the multiple accompanying emergency events, outputting multiple accompanying risk indicators. The decision-making behavior prediction tree is quantitatively labeled according to the decision-making behavior risk indicators and the multiple accompanying risk indicators, outputting a labeled decision-making behavior prediction tree. A decision-making behavior assessment report is generated by risk fitting the labeled decision-making behavior prediction tree. This method solves the technical problem in existing nuclear emergency decision-making assessment methods where dynamic risk assessment of decision-making behavior is difficult, leading to low accuracy in emergency decision-making behavior assessment and affecting the reliability of emergency decisions. By constructing a decision-making behavior prediction tree and combining it with a feedback adjustment mechanism based on the real-time number of accompanying emergency events, dynamic risk assessment and optimization of accompanying events are achieved, improving the accuracy of decision-making behavior risk assessment in nuclear emergency situations and enhancing the reliability of emergency decisions. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0022] Figure 1 A schematic diagram of a method for evaluating quantitative decision-making behavior in a nuclear emergency, provided as an embodiment of this application;

[0023] Figure 2 This is a schematic diagram of a quantitative decision-making behavior assessment system for nuclear emergency situations, provided as an embodiment of this application.

[0024] Figure labeling: 11 Accompanying event acquisition module, 12 Behavior prediction tree construction module, 13 Decision behavior risk assessment module, 14 Accompanying emergency risk assessment module, 15 Identification module, 16 Assessment report acquisition module. Detailed Implementation

[0025] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application.

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0028] This application provides a method and system for quantitatively evaluating decision-making behavior in nuclear emergency situations, such as... Figure 1 As shown, the method includes:

[0029] The system receives decision-making action plans under nuclear emergency conditions, predicts these plans, and outputs multiple accompanying emergency events. It then constructs a decision-making action prediction tree, using these accompanying emergency events as child nodes and the nuclear emergency decision-making action plans as parent nodes. Finally, it performs a risk assessment on each decision-making action plan and outputs decision-making action risk indicators.

[0030] The process involves receiving decision-making action plans for nuclear emergency situations. These plans refer to emergency response action plans made by different emergency response units or managers in response to nuclear accidents. Each plan includes how to respond to the accident, how to evacuate people, and how to mitigate the impact of the accident. Based on the received decision-making action plans, emergency events are predicted. Each decision plan may lead to a series of accompanying emergency events. Several accompanying emergency events with a high probability of transmission are obtained after screening. Further, after predicting the accompanying emergency events, a decision-making action prediction tree is constructed. The prediction tree is a structured tool that can show the relationship between decision-making actions and their accompanying emergency events. When constructing the decision-making action prediction tree, the current decision-making action plan is used as the parent node, representing the current action. The predicted accompanying emergency events are used as child nodes, indicating the secondary emergency events that may occur under different decision-making plans. For example, if the decision-making action is to initiate an emergency shutdown of the nuclear power plant, possible accompanying emergency events (child nodes) include equipment overheating, radiation spread, etc., thus completing the construction of the decision-making action prediction tree. Preferably, each accompanying emergency event can be further decomposed into smaller sub-events, forming a complex prediction tree structure. This tree structure provides a clear visual representation of the potential consequences of each decision-making action and the causal relationships between these consequences. Subsequently, a risk assessment is performed on the proposed decision-making actions, outputting risk indicators for each action.

[0031] The method provided in this application embodiment further includes: collecting historical nuclear emergency event samples; constructing an accompanying emergency event prediction model; performing simulation prediction on the historical nuclear emergency event samples and the accompanying emergency event samples based on the accompanying emergency event prediction model to obtain prediction accuracy; predicting the decision-making behavior scheme based on the accompanying emergency event prediction model, outputting an initial set of accompanying emergency events and the accompanying probability of each initial accompanying emergency event; filtering initial accompanying emergency events with a probability greater than a preset accompanying probability, and outputting the plurality of accompanying emergency events.

[0032] The decision-making action plan predicts and outputs multiple accompanying emergency events, including: collecting historical nuclear emergency event samples, which are data records of historical nuclear accidents or emergency events and their corresponding accompanying emergency events. These samples can be accident reports from governments or nuclear facilities, containing information such as the time, location, emergency response measures, and accident results. Subsequently, an accompanying emergency event prediction model is constructed to predict possible accompanying events under a specific decision-making action. Simulation predictions are performed on the historical nuclear emergency event samples and the accompanying emergency event samples based on the accompanying emergency event prediction model to obtain the prediction accuracy. The prediction accuracy is the ratio of the correct prediction result to the total prediction result. Further, the accompanying emergency event prediction model is used to predict accompanying emergency events for the decision-making action plan, outputting an initial set of accompanying emergency events. Each accompanying emergency event is assigned an accompanying probability, i.e., the likelihood of the accompanying event occurring; the higher the value, the greater the likelihood of the event occurring. For example, if the decision-making action is to shut down the reactor, the set of accompanying emergency events might include a cooling system failure with a probability of 0.6, a radiation leak with a probability of 0.4, etc. Finally, initial accompanying emergency events with a probability greater than a preset probability are selected. A preset probability threshold is set, such as 0.5, which means that only accompanying emergency events with a probability of occurrence greater than 50% are considered high-probability events. When the initial accompanying emergency event is greater than the preset probability, it is considered to have a high probability of occurrence. Then, the multiple accompanying emergency events are output according to the selection results.

[0033] The method provided in this application embodiment further includes: constructing an emergency scenario simulation model; analyzing the evolution process of the historical nuclear emergency event samples using the emergency scenario simulation model; identifying the accompanying emergency event samples of each historical nuclear emergency event sample, and the accompanying factors of each accompanying emergency event sample, wherein the accompanying factor is the probability of the accompanying event occurring. Training is performed based on the historical nuclear emergency event samples, the accompanying emergency event samples, and the accompanying factors of each accompanying emergency event sample; if the prediction accuracy reaches a preset threshold, an accompanying emergency event prediction model is output.

[0034] The method for constructing a prediction model for accompanying emergency events includes: constructing an emergency scenario simulation model, which is a mathematical model based on physical simulation, statistics, or data-driven techniques, used to simulate the evolution of scenarios after a nuclear emergency event. It considers the background conditions of the event (such as the operating status of the nuclear power plant, external environmental conditions, etc.) and predicts the possible evolution paths of the emergency event under these conditions. The evolution process of historical nuclear emergency event samples is analyzed using the emergency scenario simulation model, including horizontal evolution analysis and vertical evolution analysis. The horizontal evolution analysis simulates the diffusion scenarios of the historical nuclear emergency event samples, such as the diffusion process of a nuclear emergency event, analyzing the propagation and spread of radioactive materials, heat energy, or hazardous substances in space. The vertical evolution analysis simulates the chain scenario of the historical nuclear emergency event samples, such as a mathematical model simulating the spread behavior of radioactive materials or other emergency events in air, soil, or water. When constructing the emergency scenario simulation model, historical nuclear emergency event data is first collected and organized, including the initial state of the event, the accident handling process, the consequences, and the impact of the surrounding environment. The data should include nuclear power plant design parameters, operating history, environmental conditions, etc. A simulation engine is built using physical modeling, such as fluid dynamics, radioactive material diffusion models, and statistical modeling, to simulate the evolution of emergency events. Historical events are parameterized, with key parameters including the type, time, location, and meteorological conditions of the affected area.

[0035] Furthermore, based on the evolutionary process analysis results, accompanying emergency event samples and accompanying factors for each historical nuclear emergency event sample are identified. The accompanying factors represent the probability of the event occurring together. Finally, the neural network is trained using the historical nuclear emergency event samples, the accompanying emergency event samples, and the accompanying factors for each accompanying emergency event sample until the model's prediction accuracy reaches a preset threshold, thus efficiently outputting the accompanying emergency event prediction model.

[0036] The method provided in this application embodiment further includes: performing a partitioned simulation of the decision-making behavior scheme to obtain multiple regions; identifying the personnel protection rate and equipment protection rate of each of the multiple regions, and outputting multiple personnel protection rates and multiple equipment protection rates; and calculating and outputting a decision-making behavior risk index based on the multiple personnel protection rates and multiple equipment protection rates.

[0037] A risk assessment of the proposed decision-making action plan is conducted, outputting risk indicators. The method includes: performing a zonal simulation of the decision-making action plan to obtain multiple regions. The goal of the zonal simulation is to divide the entire accident impact area into several smaller regions to assess the risk of each region separately. Each region can be divided based on its physical location, function, or degree of impact from the accident. Dividing by specific functional areas, such as reactor areas, control room areas, cooling system areas, and surrounding community areas, is possible. After completing the zonal simulation, the personnel protection rate and equipment protection rate of each of the multiple regions are identified. The personnel protection rate refers to the proportion of people who can be protected from the accident under a specific emergency decision-making action within a certain region. The personnel protection rate is obtained by using nuclear emergency event big data to obtain the average final personnel protection rate for the corresponding emergency decision-making action in the corresponding region. The equipment protection rate refers to the proportion of equipment that can be protected from damage under a specific emergency decision-making action within a certain region. The equipment protection rate is obtained by using nuclear emergency event big data to obtain the average final equipment protection rate for the corresponding emergency decision-making action in the corresponding region. Finally, calculations are performed based on the multiple personnel protection rates and multiple equipment protection rates. During the calculation, each divided region is weighted according to its importance by professional technicians. Based on the weighting results, the multiple personnel protection rates and multiple equipment protection rates are weighted and summed to output a decision-making behavior risk index. This decision-making behavior risk index comprehensively assesses the impact of emergency decision-making behavior on personnel and equipment in nuclear emergency events, thus providing a basis for subsequent decision-making.

[0038] Risk assessments are performed on each of the multiple accompanying emergency events, resulting in multiple accompanying risk indicators. The decision-making behavior prediction tree is then quantified and labeled according to the decision-making behavior risk indicators and the multiple accompanying risk indicators, and the labeled decision-making behavior prediction tree is output. A decision-making behavior assessment report is generated by performing risk fitting on the labeled decision-making behavior prediction tree.

[0039] Risk assessments are performed on each of the multiple accompanying emergency events, outputting multiple accompanying risk indicators. Further, after obtaining all risk indicators, these indicators need to be mapped onto a prediction tree, and each node is quantified. The purpose of quantification is to assign a specific risk value to each decision-making behavior and accompanying emergency event, outputting a quantified decision-making behavior prediction tree. Finally, by performing risk fitting on the quantified decision-making behavior prediction tree, a decision-making behavior assessment report is generated. This report reflects the final risk parameter value of the decision-making behavior; a higher risk value corresponds to a higher decision risk, and vice versa. This solves the technical problem in existing nuclear emergency decision assessment methods where dynamic risk assessment of decision-making behaviors is difficult, leading to difficulty in improving the accuracy of emergency decision-making behavior assessment and affecting the reliability of emergency decisions. By constructing a decision-making behavior prediction tree and combining it with a feedback adjustment mechanism based on the real-time number of accompanying emergency events, dynamic risk assessment and optimization of accompanying events are achieved, improving the accuracy of risk assessment of decision-making behaviors in nuclear emergency events and enhancing the reliability of emergency decisions.

[0040] The method provided in this application further includes: collecting samples of accompanying emergency events and samples of accompanying loss indicators based on the event attributes of the plurality of accompanying emergency events; establishing an accompanying event risk assessment network based on the accompanying emergency event samples and the accompanying loss indicator samples; and performing a risk assessment on each accompanying emergency event based on the accompanying event risk assessment network, outputting multiple accompanying risk indicators.

[0041] The method involves conducting risk assessments on multiple accompanying emergency events and outputting multiple accompanying risk indicators. This includes: 1) Based on the event attributes of the multiple accompanying emergency events (the event attributes being the specific attribute categories of the multiple accompanying emergency events); 2) Collecting samples of accompanying emergency events identical to the accompanying emergency events, and samples of accompanying loss indicators for these accompanying emergency event samples. The accompanying emergency event samples are event record samples collected from historical nuclear emergency events that share the same attributes as the multiple accompanying emergency events. The accompanying loss indicator samples are loss data related to the accompanying emergency events. This loss data can include personal injury, property damage, environmental pollution, etc.; 3) Establishing an accompanying event risk assessment network based on the accompanying emergency event samples and the accompanying loss indicator samples. This accompanying event risk assessment network is constructed based on a neural network model. The accompanying emergency event samples and the accompanying loss indicator samples are input into the model for supervised training until the model's output accuracy meets the requirements, thus obtaining an initial accompanying event risk assessment network; 4) Adding a risk scoring calculation layer to the initial accompanying event risk assessment network to obtain the final accompanying event risk assessment network. The computational layer in the accompanying event risk assessment network is used to calculate the risk score of the accompanying loss index output by the initial accompanying event risk assessment network. Different accompanying loss indices have corresponding risk score parameters. After summing the risk score parameters corresponding to the accompanying loss indices, multiple accompanying risk indices are finally output.

[0042] The method provided in this application embodiment further includes: dividing the identified decision behavior prediction tree into hierarchical levels, outputting multi-level sub-nodes, including first-level sub-nodes, second-level sub-nodes, third-level sub-nodes, ... Nth-level sub-nodes. The multi-level sub-nodes are then labeled with levels, outputting multi-level level indicators. Using the multi-level level indicators as weights, and taking the decision behavior risk indicator and the multiple accompanying risk indicators as inputs, risk fitting is performed, outputting a risk fitting indicator. Based on the risk fitting indicator, a decision behavior assessment report is generated.

[0043] The method involves risk fitting to the labeled decision behavior prediction tree, including: dividing the tree into levels, with each level corresponding to a first-level child node, outputting multi-level child nodes, including first-level, second-level, third-level, ..., Nth-level child nodes, where N is the same as the number of levels in the prediction tree. The multi-level child nodes are then labeled with a risk level, and each level is assigned a risk level according to a preset correspondence between node levels and levels. Each risk level corresponds to a specific level indicator, and multi-level level indicators are output. Further, risk fitting is performed using the multi-level level indicators as weights, with the decision behavior risk indicator and multiple accompanying risk indicators as inputs. Specifically, the multi-level level indicators are used as weights to perform a weighted summation of the decision behavior risk indicator and the multiple accompanying risk indicators to obtain the final risk fitting index. A decision behavior assessment report is generated based on the risk fitting index.

[0044] The method provided in this application embodiment further includes: obtaining the real-time number of multiple accompanying emergency events; generating an adjustment coefficient according to the ratio of the real-time number to a preset number; and performing feedback adjustment based on the adjustment coefficient when performing risk fitting on the identified decision behavior prediction tree.

[0045] The method for outputting multiple accompanying emergency events further includes: obtaining the real-time number of multiple accompanying emergency events, where the real-time number of multiple accompanying emergency events is the output predicted number; generating an adjustment coefficient based on the ratio of the real-time number to a preset number, where the preset number is a pre-set expected number of accompanying emergency events under the current emergency situation; and performing risk fitting on the identified decision behavior prediction tree, performing feedback adjustment based on the adjustment coefficient, that is, multiplying the adjustment coefficient by the risk fitting index to optimize and adjust the risk fitting index, thereby improving the accuracy of the risk fitting index acquisition.

[0046] In the above text, refer to Figure 1 A method for evaluating quantitative decision-making behavior in a nuclear emergency, according to an embodiment of the present invention, is described in detail. Next, reference will be made to... Figure 2 This invention describes a quantitative decision-making behavior assessment system for nuclear emergency situations, according to an embodiment of the present invention.

[0047] A quantitative decision-making behavior assessment system for nuclear emergency situations, according to an embodiment of the present invention, solves the technical problem that existing nuclear emergency decision assessment methods struggle to dynamically assess the risks of decision-making behavior, leading to difficulties in improving the accuracy of emergency decision-making behavior assessment and affecting the reliability of emergency decisions. By constructing a decision-making behavior prediction tree and combining it with a feedback adjustment mechanism based on the real-time number of accompanying emergency events, dynamic risk assessment and optimization of accompanying events are achieved, thereby improving the accuracy of decision-making behavior risk assessment in nuclear emergency situations and enhancing the reliability of emergency decisions. The quantitative decision-making behavior assessment system for nuclear emergency situations includes: an accompanying event acquisition module 11, a behavior prediction tree construction module 12, a decision-making behavior risk assessment module 13, an accompanying emergency risk assessment module 14, an identification module 15, and an assessment report acquisition module 16.

[0048] The accompanying event acquisition module 11 is used to receive decision-making action plans in nuclear emergency situations, predict the decision-making action plans, and output multiple accompanying emergency events.

[0049] The behavior prediction tree construction module 12 is used to construct a decision behavior prediction tree by taking the multiple accompanying emergency events as child nodes and the decision behavior scheme under the nuclear emergency as the parent node.

[0050] The decision-making behavior risk assessment module 13 is used to assess the risk of the decision-making behavior plan and output decision-making behavior risk indicators.

[0051] The accompanying emergency risk assessment module 14 is used to conduct risk assessments on the multiple accompanying emergency events and output multiple accompanying risk indicators.

[0052] The identification module 15 is used to quantitatively identify the decision behavior prediction tree according to the decision behavior risk index and the multiple accompanying risk indicators, and output the identified decision behavior prediction tree.

[0053] The assessment report acquisition module 16 is used to generate a decision behavior assessment report by performing risk fitting on the identified decision behavior prediction tree.

[0054] The specific configuration of the accompanying event acquisition module 11 will be described in detail below. The accompanying event acquisition module 11 may further include: predicting the decision-making behavior scheme and outputting multiple accompanying emergency events, the method including: collecting historical nuclear emergency event samples; constructing an accompanying emergency event prediction model; performing simulation prediction on the historical nuclear emergency event samples and the accompanying emergency event samples based on the accompanying emergency event prediction model, and obtaining the prediction accuracy; predicting the decision-making behavior scheme based on the accompanying emergency event prediction model, outputting an initial set of accompanying emergency events, and the accompanying probability of each initial accompanying emergency event; filtering initial accompanying emergency events with a probability greater than a preset accompanying probability, and outputting them as the multiple accompanying emergency events.

[0055] The specific configuration of the accompanying event acquisition module 11 will be described in detail below. The accompanying event acquisition module 11 further includes: constructing an accompanying emergency event prediction model, the method of which includes: constructing an emergency scenario simulation model, analyzing the evolution process of the historical nuclear emergency event samples using the emergency scenario simulation model, identifying the accompanying emergency event samples of each historical nuclear emergency event sample, and the accompanying factors of each accompanying emergency event sample, wherein the accompanying factor is the magnitude of the probability of the accompanying event occurring. The model is trained based on the historical nuclear emergency event samples, the accompanying emergency event samples, and the accompanying factors of each accompanying emergency event sample. If the prediction accuracy reaches a preset threshold, the accompanying emergency event prediction model is output.

[0056] The specific configuration of the accompanying event acquisition module 11 will be described in detail below. The accompanying event acquisition module 11 may further include: analyzing the evolution process of the historical nuclear emergency event sample using the emergency scenario simulation model, including horizontal evolution analysis and vertical evolution analysis, wherein the horizontal evolution analysis is to simulate the diffusion scenario of the historical nuclear emergency event sample, and the vertical evolution analysis is to simulate the chain scenario of the historical nuclear emergency event sample.

[0057] The specific configuration of the decision-making behavior risk assessment module 13 will be described in detail below. The decision-making behavior risk assessment module 13 further includes: assessing the risk of the decision-making behavior scheme and outputting decision-making behavior risk indicators. The method includes: performing a partitioned simulation of the decision-making behavior scheme to obtain multiple regions; identifying the personnel protection rate and equipment protection rate of each of the multiple regions, and outputting multiple personnel protection rates and multiple equipment protection rates; and calculating and outputting decision-making behavior risk indicators based on the multiple personnel protection rates and multiple equipment protection rates.

[0058] The specific configuration of the accompanying emergency risk assessment module 14 will be described in detail below. The accompanying emergency risk assessment module 14 further includes: performing risk assessments on the multiple accompanying emergency events respectively, and outputting multiple accompanying risk indicators. The method includes: collecting accompanying emergency event samples and accompanying loss indicator samples based on the event attributes of the multiple accompanying emergency events; establishing an accompanying event risk assessment network based on the accompanying emergency event samples and the accompanying loss indicator samples; and performing a risk assessment on each accompanying emergency event based on the accompanying event risk assessment network, outputting multiple accompanying risk indicators.

[0059] The specific configuration of the assessment report acquisition module 16 will be described in detail below. The assessment report acquisition module 16 further includes: risk fitting of the identified decision behavior prediction tree, the method of which includes: dividing the identified decision behavior prediction tree according to its hierarchy, outputting multi-level sub-nodes, including first-level sub-nodes, second-level sub-nodes, third-level sub-nodes, ..., Nth-level sub-nodes; classifying the multi-level sub-nodes into levels, outputting multi-level level indicators; using the multi-level level indicators as weights, and the decision behavior risk indicators and the multiple accompanying risk indicators as inputs to perform risk fitting, outputting risk fitting indicators; and generating a decision behavior assessment report based on the risk fitting indicators.

[0060] The specific configuration of the assessment report acquisition module 16 will be described in detail below. The assessment report acquisition module 16 further includes: outputting multiple accompanying emergency events; the method further includes: acquiring the real-time number of the multiple accompanying emergency events. An adjustment coefficient is generated according to the ratio of the real-time number to a preset number; when performing risk fitting on the identified decision behavior prediction tree, feedback adjustment is performed based on the adjustment coefficient.

[0061] The quantitative decision-making behavior assessment system for nuclear emergency situations provided in this embodiment of the invention can execute the quantitative decision-making behavior assessment method for nuclear emergency situations provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0062] While this application makes various references to certain modules in the system according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved. In addition, the specific names of each functional unit are only for easy distinction and are not intended to limit the scope of protection of this invention.

[0063] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for evaluating quantitative decision-making behavior in nuclear emergency situations, characterized in that, The method includes: Receive decision-making action plans in the event of a nuclear emergency, predict the decision-making action plans, and output multiple accompanying emergency events; By using the multiple accompanying emergency events as child nodes and the decision-making behavior schemes under nuclear emergency conditions as parent nodes, a decision-making behavior prediction tree is constructed. A risk assessment is performed on the proposed decision-making action plan, and risk indicators for the decision-making action are output. Risk assessments are performed on each of the multiple accompanying emergency events, and multiple accompanying risk indicators are output. The decision behavior prediction tree is quantitatively labeled according to the decision behavior risk index and the multiple accompanying risk indicators, and the labeled decision behavior prediction tree is output. By performing risk fitting on the labeled decision behavior prediction tree, a decision behavior assessment report is generated. Risk assessments are conducted on the multiple accompanying emergency events, and multiple accompanying risk indicators are output. The methods include: Based on the event attributes of the multiple accompanying emergency events, collect samples of accompanying emergency events and samples of accompanying loss indicators for the accompanying emergency event samples. A risk assessment network for accompanying events is established based on the accompanying emergency event samples and the accompanying loss indicator samples; The accompanying event risk assessment network is used to assess the risk of each accompanying emergency event and output multiple accompanying risk indicators. The method involves risk fitting to the labeled decision behavior prediction tree, including: The decision behavior prediction tree after identification is divided into levels, and multi-level child nodes are output, including first-level child nodes, second-level child nodes, third-level child nodes, ... Nth-level child nodes; The multi-level child nodes are labeled with levels, and multi-level level indicators are output. Using the multi-level indicators as weights, and taking the decision-making behavior risk indicators and the multiple accompanying risk indicators as inputs, risk fitting is performed, and the risk fitting index is output. Based on the aforementioned risk fitting indicators, a decision-making behavior assessment report is generated; The method also includes outputting multiple accompanying emergency events: Obtain the real-time quantity of multiple accompanying emergency events; An adjustment coefficient is generated based on the ratio of the real-time quantity to the preset quantity. When performing risk fitting on the identified decision behavior prediction tree, feedback adjustment is performed based on the adjustment coefficient.

2. The method as described in claim 1, characterized in that, The decision-making action plan predicts and outputs multiple accompanying emergency events, including the following methods: Collect samples of historical nuclear emergency events; Construct an accompanying emergency event prediction model, and simulate and predict the historical nuclear emergency event samples and the accompanying emergency event samples based on the accompanying emergency event prediction model to obtain the prediction accuracy. The decision-making action plan is predicted based on the accompanying emergency event prediction model, and an initial set of accompanying emergency events and the accompanying probability of each initial accompanying emergency event are output. Filter out initial accompanying emergency events with a probability greater than a preset number of accompanying events, and output them as the multiple accompanying emergency events.

3. The method as described in claim 2, characterized in that, Methods for constructing prediction models for accompanying emergency events include: An emergency scenario simulation model is constructed, and the evolution process of the historical nuclear emergency event samples is analyzed using the emergency scenario simulation model. The accompanying emergency event samples of each historical nuclear emergency event sample and the accompanying factors of each accompanying emergency event sample are identified, wherein the accompanying factor is the magnitude of the probability of the event occurring together. The model is trained based on the historical nuclear emergency event samples, the accompanying emergency event samples, and the accompanying factors of each accompanying emergency event sample. If the prediction accuracy reaches a preset threshold, the accompanying emergency event prediction model is output.

4. The method as described in claim 3, characterized in that, The evolution process of the historical nuclear emergency event samples is analyzed using the emergency scenario simulation model, including horizontal evolution analysis and vertical evolution analysis. The horizontal evolution analysis is a diffusion scenario simulation of the historical nuclear emergency event samples, and the vertical evolution analysis is a chain scenario simulation of the historical nuclear emergency event samples.

5. The method as described in claim 1, characterized in that, The method for assessing the risk of the decision-making action plan and outputting decision-making risk indicators includes: The decision-making behavior scheme is simulated by partitioning to obtain multiple regions; Identify the personnel protection rate and equipment protection rate of each of the multiple regions, and output multiple personnel protection rates and multiple equipment protection rates; The decision-making behavior risk index is calculated and output based on the protection rates of the multiple personnel and the multiple equipment.

6. A quantitative decision-making behavior evaluation system for nuclear emergency situations, characterized in that, The system is used to execute a quantitative decision-making behavior assessment method in a nuclear emergency situation as described in any one of claims 1-5, including: The accompanying event acquisition module is used to receive decision-making action plans in nuclear emergency situations, predict the decision-making action plans, and output multiple accompanying emergency events; The behavior prediction tree construction module is used to construct a decision behavior prediction tree by taking the multiple accompanying emergency events as child nodes and the decision behavior plan under the nuclear emergency as the parent node. The decision-making behavior risk assessment module is used to assess the risks of the decision-making behavior plan and output decision-making behavior risk indicators. The accompanying emergency risk assessment module is used to assess the risks of the multiple accompanying emergency events and output multiple accompanying risk indicators. The identification module is used to quantitatively identify the decision behavior prediction tree according to the decision behavior risk index and the multiple accompanying risk indicators, and output the identified decision behavior prediction tree. The assessment report acquisition module is used to generate a decision behavior assessment report by performing risk fitting on the identified decision behavior prediction tree.

Images