Method and system for comprehensively evaluating risk of nuclear power station

Abstract

The invention belongs to the technical field of nuclear power station safety, and particularly relates to a method and system for comprehensively carrying out nuclear power station risk assessment. Comprising a data acquisition module, a risk assessment module, an intelligent early warning module and an emergency response module. The data acquisition module acquires data influencing the safety risk of the nuclear power station, including nuclear power station equipment operation data, radiation and environment data, personnel data and external risk data; the risk assessment module calculates a risk value through a nuclear power station risk assessment model; the intelligent early warning module performs early warning grade division and displays early warning information on a monitoring interface; and the emergency response module automatically adjusts an emergency response plan and optimizes resource configuration according to the early warning level. The system has the beneficial effects that the data acquisition and integration module acquires nuclear power station equipment operation data, radiation and environment data, personnel data and external risk data, so that factors causing potential risks of the nuclear power station are obtained, and data support is provided for comprehensive risk assessment of the nuclear power station.

Description

Technical Field

[0001] The present invention belongs to the technical field of nuclear power plant safety, and particularly relates to a method and system for comprehensively evaluating the risks of nuclear power plants. Background Art

[0002] With the continuous increase in global energy demand and the enhancement of environmental protection awareness, nuclear energy, as a clean and efficient form of energy, has been widely used globally. However, nuclear power plants face complex and changeable safety risks during operation, which not only concern the safe and stable operation of the nuclear power plants themselves, but also directly affect the surrounding environment and public safety. Therefore, nuclear power plant risk assessment and early warning technologies have emerged as the key means to ensure the safe utilization of nuclear energy.

[0003] Currently, significant progress has been made in nuclear power plant risk assessment and early warning technologies. On the one hand, by real-time monitoring various operating parameters of nuclear power plants, such as temperature, pressure, radiation level, etc., combined with advanced data acquisition and transmission technologies, comprehensive monitoring of the operating status of nuclear power plants has been achieved. On the other hand, by using big data analysis and artificial intelligence algorithms to deeply mine and process massive data, abnormal situations and potential risks can be detected in a timely manner, and early warning signals can be issued. In addition, the application of new technical means such as unmanned aerial vehicles and remote sensors has further improved the accuracy and timeliness of nuclear power plant risk assessment and early warning.

[0004] Despite the significant progress made in nuclear power plant risk assessment and early warning technologies, there are still some problems and challenges. First, the phenomenon of data islands is serious, and there are obstacles to data sharing and interoperability between different systems, resulting in the early warning system being unable to fully utilize all data resources for analysis and judgment. Second, the accuracy and reliability of risk assessment models still need to be further improved. Especially when facing complex and changeable safety risks, existing models cannot quickly and accurately predict and identify potential risks. In addition, the perfection degree of the emergency response mechanism also directly affects the actual effect of the early warning system. How to quickly initiate emergency response measures and effectively control the development of the situation within a short time is also one of the current problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for comprehensively evaluating the risks of nuclear power plants, which can effectively solve problems such as the inability of existing models to quickly and accurately predict and identify.

[0006] The technical solution of the present invention is as follows: A method for comprehensively evaluating the risks of nuclear power plants, the system applied includes a data acquisition module, a risk assessment module, an intelligent early warning module, and an emergency response module, and includes the following steps:

[0007] Step 1: The data acquisition module collects equipment operation data, radiation and environmental data, personnel data, and external risk data;

[0008] Step 2: The risk assessment module uses two objective weighting methods, the entropy weight method and the CRITIC method, to determine the importance of each data index, and then optimizes the TOPSIS method using the KL distance and the grey correlation theory to construct a risk assessment model and calculate the risk value;

[0009] Step 3: The intelligent early warning module combines historical data and real-time data of the risk value and uses machine learning methods to divide the early warning levels;

[0010] Step 4: The emergency response module automatically adjusts the emergency response plan and optimizes resource allocation according to the early warning level.

[0011] The equipment operation data includes power, temperature, pressure parameter data, and pump, valve, and indicator light status data; the radiation and environmental data includes radiation dose data and environmental parameter data; the personnel data includes personnel operation records and personnel training and education data; the external risk data includes natural disaster information and equipment risk psychological information.

[0012] The power, temperature, pressure, radiation dose, and environmental parameter data are collected by sensors, the status data is collected by a data acquisition system, the personnel data is collected by an access control system and a work log recording system, and the external risk data is integrated by data crawling technology.

[0013] The sensors include temperature sensors, pressure sensors, and radiation monitors.

[0014] Two objective weighting methods, the entropy weight method and the CRITIC method, are used to determine the importance of each data index. Specifically, the two objective weight vectors are combined through weighted averaging, and a distance function is introduced for consistency solving.

[0015] The optimization of the TOPSIS method includes: from the perspective of distance, improving the Euclidean distance in the TOPSIS method based on the KL distance, and from the perspective of geometric similarity, optimizing the TOPSIS method using grey correlation analysis.

[0016] The optimization steps of the TOPSIS method are as follows:

[0017] 1) Suppose there are p evaluation object samples and q data indicators, and the initial judgment matrix X is constructed as follows:

[0018]

[0019] where, x ijis the j-th data index value of the i-th evaluation sample;

[0020] 2) Standardize the data index values to obtain the standardized matrix Y as follows:

[0021]

[0022] where, y ij is the j-th data index value of the i-th evaluation sample after standardization;

[0023] 3) Multiply the standardized matrix Y by the comprehensive weight to obtain the weighted decision matrix Z as follows:

[0024]

[0025] where, z ij is the j-th data index value of the i-th evaluation sample in the weighted decision matrix; w j is the comprehensive weight value of the j-th data index;

[0026] 4) Standardize the weighted decision matrix Z to obtain the standardized weighted decision matrix U as follows:

[0027]

[0028] where, u ij is the j-th data index value of the i-th evaluation sample in the standardized weighted decision matrix;

[0029] 5) Determine the positive ideal solution and negative ideal solution of the standard weighted decision matrix, and for the benefit index set J + the positive ideal solution U + is the maximum value of the row vector, and the negative ideal solution U - is the minimum value of the row vector. For the cost index set J - the positive ideal solution U + is the minimum value of the row vector, and the negative ideal solution U - is the maximum value of the row vector;

[0030]

[0031] where, U j + and U j - are the positive ideal solution and negative ideal solution of the j-th data index respectively;

[0032] 6) Calculate the KL distances V i + and V i - :

[0033]

[0034] Among them, V i + and V i - are the distances between the evaluation sample and the positive ideal solution and the negative ideal solution respectively. u j + and u j - are the positive and negative ideal solutions of the j-th data index respectively;

[0035] 7) Calculate the grey correlation coefficients x ij + and x ij - :

[0036]

[0037] Among them, 、 are the maximum and minimum values of the absolute difference between the positive ideal solution of the j-th data index and the index in the standard weighted matrix respectively. α is the discrimination coefficient;

[0038] 8) Calculate the grey correlation degrees X i + and X i - :

[0039]

[0040] 9) Calculate the closeness degrees of the evaluation sample to the positive and negative ideal solutions:

[0041]

[0042] Among them, T i + 、T i - are the closeness degrees of the evaluation object to the positive ideal solution and the negative ideal solution respectively. A and B are the emphasis coefficients of the evaluation object on the distance and the curve shape respectively;

[0043] 10) Calculate the comprehensive closeness degree S i + :

[0044]

[0045] Among them, 0 ≤ S i + ≤ 1.

[0046] The division of the warning levels is specifically as follows: Select a machine learning method, use the historical data of risk values as the training set to train the model, and repeatedly adjust the model parameters so that the model can learn the distribution pattern of risk values and the boundary characteristics between different risks, thereby constructing the division rules of the warning levels.

[0047] The emergency response plan selects corresponding measures according to four different warning levels: minor, critical, severe, and catastrophic.

[0048] A system for comprehensively evaluating the risks of a nuclear power plant, including a data acquisition module, a risk assessment module, an intelligent warning module, and an emergency response module;

[0049] The data acquisition module acquires data affecting the safety risks of the nuclear power plant, including the operation data of nuclear power plant equipment, radiation and environmental data, personnel data, and external risk data;

[0050] The risk assessment module calculates the risk value through the nuclear power plant risk assessment model;

[0051] The intelligent warning module conducts the division of warning levels and displays the warning information on the monitoring interface;

[0052] The emergency response module automatically adjusts the emergency response plan according to the warning level and optimizes the resource allocation.

[0053] The beneficial effects of the present invention are as follows:

[0054] (1) The data acquisition and integration module in the present invention acquires the operation data of nuclear power plant equipment, radiation and environmental data, personnel data, and external risk data, thereby obtaining the factors leading to the potential risks of the nuclear power plant and providing data support for the comprehensive risk assessment of the nuclear power plant.

[0055] (2) To avoid the influence of human subjective factors, the present invention uses a combined objective weighting method to determine the importance of each data index, and uses the grey relational theory and KL distance (Kullback - Leibler Divergence) to improve the traditional TOPSIS method (Technique for Order Preference by Similarity to an Ideal Solution). The combination of the two can conduct a more comprehensive risk assessment.

[0056] (3) The intelligent warning module realizes the automation and precision of risk warning and the rapidity and intelligence of emergency response by combining machine learning. The emergency response module can select corresponding measures according to four different warning levels: minor, critical, severe, and catastrophic. Compared with the prior art, through the above two modules, not only the accuracy and timeliness of risk identification are improved, but also the configuration and scheduling of emergency resources are optimized, continuously enhancing the overall response ability and robustness of the system. Description of the Drawings

[0057] Figure 1 Block diagram of a system structure for comprehensively conducting risk assessment of nuclear power plants provided by the present invention;

[0058] Figure 2 Flowchart for constructing the risk assessment module in the present invention. Specific implementation manners

[0059] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0060] The present invention aims to propose a method and system for comprehensively conducting risk assessment of nuclear power plants to solve problems such as data islands, insufficient accuracy of assessment models, and lag in emergency response in the prior art.

[0061] As Figure 1 shown, a system for comprehensively conducting risk assessment of nuclear power plants includes a data acquisition module, a risk assessment module, an intelligent early warning module, and an emergency response module.

[0062] Data acquisition module: Collect data affecting the safety risks of nuclear power plants, including nuclear power plant equipment operation data, radiation and environmental data, personnel data, and external risk data; the risk assessment module calculates the risk value through the nuclear power plant risk assessment model; the intelligent early warning module conducts early warning level division and displays the early warning information on the monitoring interface; the emergency response module automatically adjusts the emergency response plan and optimizes resource allocation according to the early warning level.

[0063] Combined with Figure 1 and Figure 2 , a method for risk assessment based on a system for comprehensively conducting risk assessment of nuclear power plants includes the following steps:

[0064] Step 1: Data acquisition

[0065] Collect data affecting the safety risks of nuclear power plants, including nuclear power plant equipment operation data, radiation and environmental data, personnel data, and external risk data. The equipment operation data includes power, temperature, pressure parameter data, and pump, valve, and indicator light status data; the radiation and environmental data includes radiation dose data and environmental parameter data; the personnel data includes personnel operation records and personnel training and education data; the external risk data includes natural disaster information and equipment risk psychological information. Among them, power, temperature, pressure, radiation dose, and environmental parameter data are collected through sensors, status data is collected through a data acquisition system, personnel data is collected through an access control system and a work log recording system, and external risk data is integrated through data crawling technology. The sensors include temperature sensors, pressure sensors, radiation monitors, etc.

[0066] Use methods such as regular expressions and data filtering to remove invalid and incorrect data from the collected data. Then, through data conversion tools, unify data from different sources and in different formats into a semi-structured format, and apply the hash algorithm to remove duplicates and the interpolation method to fill in missing data, thereby realizing the normalization processing of the data to ensure that the data is analyzed subsequently on the same scale.

[0067] Step 2: Risk assessment

[0068] To avoid the influence of human subjective factors, this invention uses two objective weighting methods, the entropy weight method and the CRITIC method, to determine the importance of each data indicator, and uses the grey relational theory and the KL distance (Kullback - Leibler Divergence) to improve the traditional TOPSIS method (technique for order preference by similarity to an ideal solution). The grey relational theory focuses on the correlation between data indicators, and the KL distance focuses on the data distribution difference. The combination of the two can conduct a more comprehensive risk assessment.

[0069] (1) Determine the comprehensive weight of each data indicator

[0070] Currently, most combined weighting methods choose the multiplication combination or the linear weighting method, so the rationality and reliability of the selection of the weight preference coefficient and its calculation method still need to be further verified. To improve this situation, a distance function is introduced, and based on this, an equation for the difference degree relationship between objective weight preference coefficients is constructed, and a more accurate combined weight is obtained by solving this equation.

[0071] Suppose there are p evaluation samples and q data indicators. If the weight vector obtained by the entropy weight method is w e , and the weight vector obtained by the CRITIC method is w c , at the same time, the weight distribution coefficients of these two methods are k1 and k2 respectively, and the comprehensive weight w j can combine the two objective weight vectors by weighted average, and the specific formula is as follows:

[0072] (1)

[0073] Among them, w ej , w cj are the weight values calculated by the entropy weight method and the CRITIC method for the jth data indicator respectively.

[0074] To quantify the difference between the preference coefficients generated by the entropy weight method and the CRITIC method when determining the index weights, the Euclidean distance function D is introduced as a tool for consistency solving. The calculation formula is as follows:

[0075] (2)

[0076] By solving the simultaneous equations (1) and (2), the comprehensive weight w can be obtained. j .

[0077] (2) Construct a nuclear power plant risk assessment model

[0078] Optimize and upgrade the traditional TOPSIS method using grey relational analysis and KL distance. The calculation steps are as follows:

[0079] 1) Suppose there are p evaluation samples and q data indicators. The initial evaluation matrix X is constructed as follows:

[0080] (3)

[0081] where, x ij is the value of the j-th data indicator of the i-th evaluation sample.

[0082] 2) Standardize the data indicator values to obtain the standardized matrix Y as follows:

[0083] (4)

[0084] where, y ij is the value of the j-th data indicator of the i-th evaluation sample after standardization.

[0085] 3) Multiply the standardized matrix Y by the comprehensive weight to obtain the weighted decision matrix Z as follows:

[0086] (5)

[0087] where, z ij is the value of the j-th data indicator of the i-th evaluation sample in the weighted decision matrix; w j is the comprehensive weight value of the j-th data indicator.

[0088] 4) Standardize the weighted decision matrix Z to obtain the standardized weighted decision matrix U as follows:

[0089] (6)

[0090] where, u ij is the value of the j-th data indicator of the i-th evaluation sample in the standardized weighted decision matrix.

[0091] 5) Determine the positive ideal solution and negative ideal solution of the standard weighted decision matrix. The positive ideal solution U + of the benefit index set J is the maximum value of the row vector, and the negative ideal solution U + is the minimum value of the row vector. The positive ideal solution U - of the cost index set J - is the minimum value of the row vector, and the positive ideal solution U +is the minimum value of the row vector, and the negative ideal solution U - is the maximum value of the row vector.

[0092] (7)

[0093] where U j + and U j - are the positive ideal solution and the negative ideal solution of the j-th data index respectively.

[0094] 6) Calculate the KL distances V i + and V i - from the evaluation object to the positive and negative ideal solutions:

[0095] (8)

[0096] where V i + and V i - are the distances between the evaluation sample and the positive ideal solution, and the negative ideal solution respectively; u j + and u j - are the positive and negative ideal solutions of the j-th data index respectively.

[0097] 7) Calculate the grey correlation coefficients x ij + and x ij -

[0098] (9)

[0099] where and are the maximum and minimum values of the absolute difference between the positive ideal solution of the j-th data index and the index in the standard weighted matrix respectively; α is the resolution coefficient, 0 ≤ α ≤ 1, usually taken as 0.5.

[0100] 8) Calculate the grey correlation degrees X i + and X i - :

[0101] (10)

[0102] 9) Calculate the closeness of the evaluation sample to the positive and negative ideal solutions:

[0103] (11)

[0104] Among them, T i + and T i - are the closeness degrees of the evaluation object to the positive ideal solution and the negative ideal solution respectively; A and B are the weighting coefficients of the evaluation object for distance and curve shape respectively, and A = B = 0.5 is taken.

[0105] 10) Calculate the comprehensive closeness degree S of the evaluation sample to the positive and negative ideal solutions i + :

[0106] (12)

[0107] Among them, 0 ≤ S i + ≤ 1, and the closer S i + is to 1, the higher the possibility of potential risk.

[0108] The process from formula (3) to formula (12) is the construction process of the risk assessment model. The calculated comprehensive closeness degree results of the evaluation sample to the positive and negative ideal solutions are the risk values of the evaluation sample; compared with the original TOPSIS method, the risk assessment model of the present invention improves the Euclidean distance in the original TOPSIS method from the distance perspective based on relative entropy (KL distance), as shown in formula (8), and optimizes the original TOPSIS method from the geometric similarity perspective by using grey relational analysis, as shown in formulas (9) to (10).

[0109] Step 3: Intelligent early warning

[0110] First, select a machine learning method such as the decision tree classification algorithm, and use most of the historical data of the risk value as the training set to train the model. By repeatedly adjusting the model parameters, the model can learn the distribution pattern of the risk value and the boundary characteristics between different risks, so as to construct the classification rules including four early warning levels: slight, critical, severe, and catastrophic. Then, use the remaining historical data to verify and optimize the trained model to ensure the accuracy and stability of the model. Finally, input the real-time data of the risk value into the trained and optimized model, and the model can classify the risk value into the corresponding level according to the learned classification rules, so as to realize the classification of the risk value level based on machine learning. At the same time, use technologies such as GIS and 3D visualization to intuitively display early warning information such as early warning levels, risk locations, and risk types on the monitoring interface.

[0111] This method can automatically identify changes in warning levels, improve the accuracy and timeliness of warnings, and formulate corresponding response measures and warning mechanisms in advance for the four warning levels of minor, critical, severe, and catastrophic.

[0112] Step 4: Emergency Response

[0113] According to the warning level, the emergency response plan is automatically adjusted to optimize resource allocation to ensure that response measures can be taken quickly and effectively when risks occur. Specifically:

[0114] (1) When the risk level is low, the system automatically dispatches professional inspectors to the risk area through SMS, email, APP push, etc., and uses portable detection equipment to conduct more detailed on-site inspections. The inspection results and treatment suggestions are promptly fed back.

[0115] (2) When the risk level is critical, initiate the emergency material allocation procedure to ensure that protective equipment, emergency repair tools, spare parts and other materials can be delivered to the site quickly, notify the maintenance team to assemble quickly and be on standby to carry out precise maintenance work in a timely manner, and notify relevant departments to make preparations for safety protection and personnel evacuation in the surrounding areas to prevent further escalation of risks.

[0116] (3) When the risk level is serious, organize the surrounding personnel to evacuate to a safe area in an orderly manner through the broadcasting system, emergency indicator lights, etc. Notify the emergency command center to take over the control of all systems, and notify and coordinate external rescue forces (such as firefighting, medical, etc.) to rush to the scene and stand by. At the same time, activate the accident investigation team to quickly determine the cause of the accident and formulate a targeted rescue plan.

[0117] (4) In the case of catastrophic risk, in addition to the above measures, the highest level emergency response mechanism will be activated to achieve seamless information connection with relevant departments and fully accept unified command and coordination. Combined with drones, satellite remote sensing and other technologies, the entire accident site will be monitored in all directions to provide real-time and accurate on-site information for emergency rescue. Actively cooperate in subsequent environmental monitoring, pollution control and social stability maintenance to minimize the losses and impacts caused by the accident on personnel, the environment and society. After the risks are effectively controlled, collect and organize professional personnel to conduct a comprehensive review and evaluation of the cause of the accident and the emergency response process, and provide detailed data and experience support for the optimization of subsequent emergency response plans.

[0118] The embodiments are preferred implementations of the present invention, but the present invention is not limited to the above-mentioned implementations. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essential content of the present invention belong to the protection scope of the present invention.

Claims

1. A method for comprehensively conducting risk assessment of nuclear power plants, the system applied by which includes a data acquisition module, a risk assessment module, an intelligent early warning module, and an emergency response module, characterized in that, It includes the following steps: Step 1: The data acquisition module acquires equipment operation data, radiation and environmental data, personnel data, and external risk data; Step 2: The risk assessment module uses two objective weighting methods, the entropy weight method and the CRITIC method, to determine the importance of each data indicator, and then optimizes the TOPSIS method using the KL distance and grey correlation theory to construct a risk assessment model and calculate the risk value; Step 3: The intelligent warning module combines the historical data and real-time data of the risk value and uses machine learning methods to divide the warning levels; Step 4: The emergency response module automatically adjusts the emergency response plan and optimizes the resource allocation according to the warning level.

2. The method according to claim 1, characterized in that: The equipment operation data includes power, temperature, pressure parameter data, and pump, valve, indicator light status data; the radiation and environmental data includes radiation dose data and environmental parameter data; the personnel data includes personnel operation records and personnel training and education data; the external risk data includes natural disaster information and equipment risk psychological information.

3. The method according to claim 2, characterized in that: The power, temperature, pressure, radiation dose, and environmental parameter data are collected by sensors, the status data is collected by a data acquisition system, the personnel data is collected by an access control system and a work log recording system, and the external risk data is integrated by data crawling technology.

4. The method according to claim 3, wherein: The sensors include temperature sensors, pressure sensors, and radiation monitors.

5. The method according to claim 1, wherein: Two objective weighting methods, the entropy weight method and the CRITIC method, are used to determine the importance of each data indicator. Specifically, the two objective weight vectors are combined through weighted averaging, and a distance function is introduced for consistency solution.

6. The method according to claim 1, characterized in that: The optimization of the TOPSIS method includes improving the Euclidean distance in the TOPSIS method based on the KL distance from the distance perspective, and optimizing the TOPSIS method using grey correlation analysis from the geometric similarity perspective.

7. The method according to claim 6, wherein: The optimization steps of the TOPSIS method are as follows: 1) Suppose there are p evaluation object samples and q data indicators, and the initial judgment matrix X is constructed as follows: where x ij is the j-th data metric value of the i-th evaluation sample; 2) Standardize the data indicator values to obtain the standardized matrix Y as follows: where y ij is the j-th data metric value of the i-th evaluation sample after standardization; 3) Multiply the standardized matrix Y by the comprehensive weight to obtain the weighted decision matrix Z as follows: where z ij is the j-th data index value of the i-th evaluation sample in the weighted decision matrix; w j is the comprehensive weight value of the j-th data index; 4) Standardize the weighted decision matrix Z to obtain the standardized weighted decision matrix U as follows: where, u ij is the j-th data index value of the i-th evaluation sample in the standardized weighted decision matrix; 5) Determine the positive ideal solution and negative ideal solution of the standard weighted decision matrix, and the positive ideal solution U of the benefit index set J + is the maximum value of the row vector, and the negative ideal solution U + is the minimum value of the row vector. For the cost index set J - the positive ideal solution U - is the minimum value of the row vector, and the negative ideal solution U + is the maximum value of the row vector; - ​ Among them, U j + and U j - are the positive ideal solution and the negative ideal solution of the j-th data index respectively; 6) Calculate the KL distances V from the evaluation object to the positive and negative ideal solutions i + and V i - : Among them, V i + and V i - are the distances between the evaluation sample and the positive ideal solution and the negative ideal solution respectively, and u j + and u j - are the positive and negative ideal solutions of the j-th data index respectively; 7) Calculate the grey correlation coefficients x of the evaluation samples to the positive and negative ideal solutions ij + and x ij - : Among them, and are the maximum and minimum values of the absolute difference between the positive ideal solution of the j-th data index and the index in the standard weighted matrix, respectively, and α is the discrimination coefficient; 8) Calculate the grey relational degrees X of the evaluation samples to the positive and negative ideal solutions i + and X i - : 9) Calculate the closeness of the evaluation samples to the positive and negative ideal solutions: Among them, T i + and T i - are the closeness degrees of the evaluation object to the positive ideal solution and the negative ideal solution respectively, and A and B are the weighting coefficients of the evaluation object for distance and curve shape respectively; 10) Calculate the comprehensive closeness degree S of the evaluation sample to the positive and negative ideal solutions i + : where 0 ≤ S i + ≤ 1.

8. The method according to claim 1, wherein: The division of the warning levels is specifically to select a machine learning method, use the historical data of the risk value as the training set to train the model, and repeatedly adjust the model parameters so that the model can learn the distribution pattern of the risk value and the boundary characteristics between different risks, thereby constructing the division rules of the warning levels.

9. The method according to claim 1, wherein: The emergency response plan selects corresponding measures according to four different warning levels: minor, critical, severe, and disaster.

10. A system applying the method according to claim 1, characterized in that: It includes a data acquisition module, a risk assessment module, an intelligent warning module, and an emergency response module; The data acquisition module: acquires data affecting the safety risks of nuclear power plants, including nuclear power plant equipment operation data, radiation and environmental data, personnel data, and external risk data; The risk assessment module: calculates the risk value through the nuclear power plant risk assessment model; The intelligent early warning module: classifies the early warning levels and displays the early warning information on the monitoring interface; The emergency response module: automatically adjusts the emergency response plan according to the early warning levels and optimizes the resource allocation.

Images