CN115798760A - Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant - Google Patents
Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant Download PDFInfo
- Publication number
- CN115798760A CN115798760A CN202211224907.7A CN202211224907A CN115798760A CN 115798760 A CN115798760 A CN 115798760A CN 202211224907 A CN202211224907 A CN 202211224907A CN 115798760 A CN115798760 A CN 115798760A
- Authority
- CN
- China
- Prior art keywords
- accident
- power plant
- tracking
- nuclear
- diagnosis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Monitoring And Testing Of Nuclear Reactors (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention belongs to the technical field of nuclear safety and nuclear emergency, and particularly relates to an accident online diagnosis tracking and rapid prediction system and method suitable for a nuclear power plant. The system comprises a nuclear accident diagnosis module, a nuclear accident tracking module and a nuclear accident rapid prediction and intervention measure evaluation module. The beneficial effects are that: 1) The tracking of the state of the nuclear power unit and the diagnosis of accident reasons can be completed based on a small number of monitoring signals of the nuclear power unit; 2) The tracking and diagnosis of the accident state can be completed within 300 seconds after the data is received, and the accuracy is over 95 percent; 3) The diagnosis information of the accident is comprehensive, and comprises accident types (LOCA, SBO, MSLB and the like), accident severity, accident occurrence areas and the like; 4) Taking a unit monitoring signal as an input, and outputting a series of unmonitorable signals by using a circulating neural network, wherein the series of unmonitorable signals comprise a reactor core state, a loop node state and a containment node state; the fast-time multiplying power of the fast prediction can reach more than 40 times of the overall fast-time effect.
Description
Technical Field
The invention belongs to the technical field of nuclear safety and nuclear emergency, and particularly relates to an accident online diagnosis, tracking, prediction and evaluation system and method for a third-generation advanced passive nuclear power plant.
Background
The nuclear power plant has a complex structure and numerous related devices, and in order to timely and comprehensively master the state of the nuclear power plant, each device is provided with a plurality of sensors for real-time monitoring, so that once the operation of the nuclear power plant is abnormal, particularly when the nuclear power plant is developed under the condition of a serious accident, a power plant operator and an emergency response technician position abnormal information from mass monitoring data and alarm signals, and then make a correct decision. Even in a severe accident stage, the loss of power supply and the large-area damage or failure of instruments are often accompanied, and at the moment, the power plant state monitoring information provided by the nuclear power plant is incomplete and accurate, so that the difficulty of correct decision making is further aggravated. If the wrong operation is taken due to the misjudgment of personnel or other people are in danger, the accident is likely to be worsened sharply. Therefore, the relevant personnel can have great working pressure and psychological pressure under the condition. Under the condition that a nuclear power plant has a serious accident, the root cause and the current operation state of the accident of the power plant need to be diagnosed according to limited credible operation data, the accident evolution process and the intervention measures for effectively relieving the accident development are analyzed, and finally the power plant is recovered to a controllable state, which is a great challenge for nuclear emergency response personnel. It is urgently needed that a related support system can diagnose the root cause of an accident, track and acquire the running state of a unit, evaluate the implementation effect of different intervention measures, and provide powerful technical support and auxiliary decision for emergency treatment of a nuclear accident.
In the existing systems, some systems for diagnosing, evaluating and predicting nuclear power plant accidents also appear, which generally include the following systems:
1) The system provides a method for diagnosing the root cause of a serious accident of a nuclear power plant, and mainly matches related characteristic parameters in a sample library of serious accident sequences established in advance, and once the characteristic parameters are matched with a certain accident or a plurality of accident sequences in the sample library, the possible probability of the accident sequence or a plurality of accident sequences is output as a diagnosis result, and the accident is used as the input of a prediction module to predict the development of the accident and the effect of intervention measures.
However, this system has some serious drawbacks, in particular:
1. the diagnosis result depends on the sample library seriously, the serious accident is the superposition of multiple accidents, the time of the initial accident and the time of the superposition accident are related, and the sample library is difficult to enumerate all accident sequences;
2. the judgment of the accident root cause is based on the matching of the characteristic parameters and the sample library, the temperature rise and melting of the reactor core are basically consistent for the later phenomena of the serious accident, the accident root cause possibly given by the matching mode of the characteristic parameters is various, the accurate diagnosis result is difficult to be given, and the support for the emergency response is limited;
3. the diagnosis based on the root cause is used as the input of a prediction calculation module, so that the progress of accidents is difficult to predict accurately, and because the prediction of serious accidents is not only related to accident sequences, but also related to the occurrence moments of various accidents, the prediction based on the accident sequences can not reflect the real evolution of the nuclear power accident state correctly;
4. considering that the third generation nuclear power plant is provided with passive special safety facilities, factors influencing the accident progress are more complex, and the difficulty of accurately predicting the accident progress is further improved.
Disclosure of Invention
The invention aims to provide an accident online diagnosis tracking and rapid prediction system and method suitable for a nuclear power plant, which can help nuclear emergency response personnel to rapidly acquire the overall state of an accident unit and the root cause of an accident, initialize a prediction calculation module based on the acquired unit operation state and the accident root cause, rapidly predict the development of the accident and evaluate the effect of an intervention scheme, provide technical support for nuclear emergency response, and reduce the accident consequence to the maximum extent.
The technical scheme of the invention is as follows: an accident online diagnosis, tracking and rapid prediction system suitable for a nuclear power plant comprises a nuclear accident diagnosis module, a nuclear accident tracking module and a nuclear accident rapid prediction and intervention measure evaluation module.
The nuclear accident diagnosis module comprises the following components:
(1) Read and process power plant data, handle a small amount of signals of gathering from the power plant, include: reading in power plant rough data, checking whether power plant signals are in a reasonable range, checking the consistency of multiple signal channels of the power plant, finishing the combination processing of parameter signal averaging, identifying damaged steam generator loops, combining high-pressure and low-pressure safety injection data, and finishing signal unit conversion matched with a bottom layer analysis program;
(2) And (3) overall state identification: identifying the current operating state of the power plant according to the input power plant signals, and judging the severity of the accident, wherein the severity of the accident mainly comprises reactor core submergence, partial reactor core exposure, complete reactor core exposure, reactor core damage, pressure vessel failure and containment failure;
(3) The method comprises the steps of identifying the accident type in detail, identifying signs and types of accidents, identifying events which occur in previous and current power plant data sampling intervals, and setting an event sequence and boundary conditions of a thermal hydraulic calculation engine by evaluating mass-energy change rates of a steam generator, a main system and a containment and checking whether the containment fails.
The nuclear accident tracking module is used for:
(1) Establishing and operating a thermal hydraulic program tracking process, calling and controlling the calculation of a thermal hydraulic calculation engine, and finishing the calculation of an accident state based on input data, a power plant state and an accident category identification result as input;
(2) Adjusting boundary conditions of a thermal hydraulic program, and if the tracking of the state of the power plant is seriously deviated, initializing calculation; if the state of the power plant is well tracked and no typical event response occurs, fine adjustment or no adjustment is carried out on system parameters within a small range according to real-time data of the power plant;
(3) And comparing the tracking result with the power plant data, comparing the power plant data with the calculation value of the thermodynamic calculation engine, checking the comparison result to determine whether the tracking module needs to be reinitialized or not, if so, performing assignment of an analysis program according to the current result.
The rapid nuclear accident prediction and intervention measure evaluation module comprises:
(1) Acquiring the current unit state calculated by the tracking module as an initial condition of a thermodynamic calculation engine in the prediction module;
(2) By means of the characteristic that a thermodynamic and hydraulic calculation engine in a prediction module can calculate time by time, the future development of the accident is rapidly predicted;
(3) Evaluating the implemented intervention measures based on the severe accident management guide rule used by the nuclear power plant;
(4) And the prediction of a plurality of servers is supported, and the comparative evaluation of the accident evolution conditions under different intervention schemes and assumed conditions is realized.
An on-line diagnosis, tracking and rapid prediction method suitable for nuclear power plant accidents comprises the following steps:
step 1: the nuclear accident diagnosis module is driven by using limited power plant signals as data of a system, identifies signs and types of accidents, identifies events occurring in previous and current power plant data sampling intervals, evaluates the area and position of LOCA, checks whether a containment vessel fails or not and the like as diagnosis results;
step 2: the nuclear accident tracking module is used for correcting the event sequence and the boundary condition of the calculation analysis engine in real time according to the power plant signal and the diagnosis result and carrying out online tracking on the running state of the accident power plant;
and 3, step 3: the nuclear accident rapid prediction and intervention measure evaluation module takes the diagnosis and tracking result as a calculation initial condition, utilizes the MAAP (severe accident analysis software) to predict the state of a future power plant in advance, realizes early warning on a key accident event and supports intervention measure insertion prediction calculation;
and 4, step 4: by utilizing a circulating neural network in a deep learning algorithm, excavating implicit information carried by a power plant signal to supplement a diagnosis model;
and 5: synchronously comparing the MAAP calculation result with the power plant signal, and confirming whether the tracking calculation is consistent with the state of the unit;
and 6: the method supports the introduction of the potential intervention measures into the prediction calculation in the nuclear power plant serious accident management guide rule, combines the positive evaluation, the negative evaluation and the long-term concern items of the intervention measures with the prediction calculation result, realizes the dynamic evaluation of the implementation effect of the intervention measures, and supports the simultaneous analysis and comparison of four different intervention schemes.
The step 1 further comprises:
and (4) preprocessing the data of the power plant, reading limited monitorable signals of the power plant, carrying out validity check on the read-in power plant signals, identifying and processing abnormal points, and processing the subsequent analysis and calculation of a data driving system.
The step 2 further comprises the following steps:
and the nuclear accident tracking module adopts a severe accident analysis software MAAP and builds a physical model of the nuclear power plant.
The step 3 further comprises:
the nuclear accident rapid prediction and intervention measure evaluation module also adopts a severe accident analysis software MAAP and builds a physical model of the nuclear power plant;
by means of calculating and analyzing the super real-time performance of the engine, predicting the subsequent development of the accident, and predicting the occurrence time of important events, namely predicting the occurrence time of the important events (such as core melting, pressure vessel lower end socket failure and containment failure);
and aiming at 5 different assumptions of optimal estimation, conservative estimation, optimistic estimation, conservative estimation and optimal estimation, and optimistic estimation, a multi-thread parallel computing mode is adopted.
The step 5 further comprises the following steps:
if the state tracking consistency is good and no typical event response occurs, performing small-range fine adjustment or no adjustment on the MAAP according to the current power plant signal and the diagnosis result, otherwise, performing reinitialization on the MAAP.
The step 6 further comprises:
and generating a corresponding prediction report according to the comparison and analysis of parameters among different accident processes and the comparison and analysis of the accident consequences at different moments and under different intervention schemes to judge the effectiveness of the implementation of intervention measures.
The invention has the beneficial effects that: 1) The state tracking and accident reason diagnosis of the nuclear power unit can be completed based on a small number of monitoring signals (120 are needed for the third-generation passive pressurized water reactor) of the nuclear power unit;
2) The tracking and diagnosis of the accident state can be completed within 300 seconds after the data are received, and the accuracy is over 95 percent;
3) The diagnosis information of the accident is comprehensive, and comprises accident types (LOCA, SBO, MSLB and the like), accident severity, accident occurrence areas and the like;
4) The method comprises the following steps of taking a unit monitoring signal as an input, and outputting a series of unmonitorable signals by utilizing a cyclic neural network, wherein the outputs comprise a reactor core state, a loop node state, a containment node state and the like;
5) The rapid prediction can predict the effects of four different intervention schemes at the same time, each intervention scheme can simultaneously give out the accident processes of 5 different hypothesis models including optimal estimation, conservation, optimism, conservation and optimal estimation, and optimism and optimal estimation, namely 20 evolution modes of an accident can be provided at most at one time;
6) The fast-time multiplying power of the fast prediction can reach more than 40 times of the overall fast-time effect.
Drawings
FIG. 1 is a flow diagram of a nuclear accident online diagnosis and status tracking module;
FIG. 2 is a flow diagram of a module for rapid prediction of nuclear accidents and intervention evaluation.
Detailed Description
The invention is described in further detail below with reference to the figures and the embodiments.
In the third generation advanced passive nuclear power plant accident online diagnosis tracking and rapid prediction system, modeling is carried out aiming at various passive special safety facilities, and the simulation calculation of the serious accident of the third generation passive nuclear power plant is realized. The method can diagnose accident symptoms, accident intervals, accident root causes, accident sequences and equipment states based on a small amount of collected unit operation data, track the operation state of the current unit through a bottom layer thermal hydraulic analysis program and an algorithm for predicting and correcting a calculation boundary, and realize simulation calculation and real-time tracking of the nuclear power plant accident. Fuzzy recognition and deep learning algorithms are introduced, and the accident diagnosis accuracy and tracking consistency of the third-generation passive nuclear power plant are greatly improved. Meanwhile, the calculation result of accident diagnosis and tracking can be used as an initial condition, the future unit state can be predicted and calculated through a bottom layer thermal hydraulic analysis program, and the frame structure is shown in figure 1.
An accident online diagnosis, tracking and rapid prediction system suitable for a nuclear power plant comprises: the system comprises a nuclear accident diagnosis module, a nuclear accident tracking module and a nuclear accident rapid prediction and intervention measure evaluation module.
The nuclear accident diagnosis module is used for:
(1) Read and process power plant data, handle a small amount of signals of gathering from the power plant, include: reading in power plant rough data, checking whether power plant signals are in a reasonable range, checking the consistency of multiple signal channels of the power plant, finishing the combination processing of parameter signal averaging, identifying damaged steam generator loops, combining high-pressure and low-pressure safety injection data, and finishing signal unit conversion matched with a bottom layer analysis program;
(2) And (3) overall state identification: identifying the current operating state of the power plant according to an input power plant signal, and judging the severity of an accident, wherein the severity of the accident mainly comprises reactor core submergence, partial reactor core uncovering, complete reactor core uncovering, reactor core damage, pressure vessel failure and containment failure;
(3) The method comprises the steps of identifying accident types in detail, identifying signs and types of accidents, identifying events occurring in previous and current power plant data sampling intervals, and setting event sequences and boundary conditions of a thermal hydraulic calculation engine by evaluating mass-energy change rates of a steam generator, a main system and a containment and checking whether the containment fails.
A nuclear accident tracking module:
(1) Establishing and operating a thermal hydraulic program tracking process, calling and controlling the calculation of a thermal hydraulic calculation engine, and finishing the calculation of an accident state based on input data, a power plant state and an accident category identification result as input;
(2) Adjusting boundary conditions of a thermal hydraulic program, and if the tracking of the state of the power plant is seriously deviated, initializing calculation; if the state of the power plant is tracked well and no typical event response occurs, fine adjustment or no adjustment is carried out on the system parameters in a small range according to the real-time data of the power plant;
(3) And comparing the tracking result with the power plant data, comparing the power plant data with the calculation value of the thermodynamic calculation engine, checking the comparison result to determine whether the tracking module needs to be reinitialized or not, if so, performing assignment of an analysis program according to the current result.
The nuclear accident rapid prediction and intervention measure evaluation module comprises:
(1) Acquiring the current unit state calculated by the tracking module as an initial condition of a thermodynamic calculation engine in the prediction module;
(2) By means of the characteristic that a thermodynamic and hydraulic calculation engine in a prediction module can perform time doubling calculation, the future development of the accident is rapidly predicted;
(3) Evaluating the implemented intervention measures based on a Severe Accident Management Guide (SAMG) used by the nuclear power plant;
(4) And the prediction of a plurality of servers is supported, and the comparative evaluation of the accident evolution conditions under different intervention schemes and assumed conditions is realized.
The accident online diagnosis tracking and rapid prediction method suitable for the nuclear power plant comprises the following steps:
step 1: the nuclear accident diagnosis module is driven by using limited power plant signals as data of a system, identifies signs and types of accidents, identifies events occurring in previous and current power plant data sampling intervals, evaluates LOCA area and position (if a breach sequence), checks whether a containment vessel fails or not and the like as diagnosis results;
the step 1 further comprises:
preprocessing power plant data, reading limited power plant monitorable signals, performing validity check on the read power plant signals, identifying and processing abnormal points, and driving a system to perform subsequent analysis and calculation by the processed data;
the diagnosis model takes an expert knowledge base method as a core and combines a fault tree and a fuzzy recognition method.
And 2, step: the nuclear accident tracking module is used for correcting an event sequence and boundary conditions of a calculation analysis engine (MAAP program) in real time according to the power plant signals and the diagnosis result and carrying out online tracking on the operation state of the accident power plant;
the step 2 further comprises the following steps:
the nuclear accident tracking module adopts the international widely applied serious accident analysis software MAAP and builds a physical model aiming at the characteristics of the third-generation advanced passive nuclear power plant.
And step 3: the nuclear accident rapid prediction and intervention measure evaluation module takes the diagnosis and tracking result as the initial calculation condition, utilizes the MAAP time-doubled calculation characteristic of the serious accident analysis software to predict the state of the power plant in the future in advance, realizes the early warning of the key accident event and supports the intervention measure insertion prediction calculation.
The third step further comprises:
the nuclear accident rapid prediction and intervention measure evaluation module also adopts the international widely-applied serious accident analysis software MAAP, and builds a physical model aiming at the characteristics of the third-generation advanced passive nuclear power plant;
by means of calculating and analyzing the super real-time performance of the engine, predicting the subsequent development of the accident, and predicting the occurrence time of important events, namely predicting the occurrence time of the important events (such as core melting, pressure vessel lower head failure and containment failure);
and aiming at 5 different assumptions of optimal estimation, conservation and optimism, and between conservation and optimal estimation and between optimism and optimal estimation, a multi-thread parallel computing mode is adopted, and the uncertainty of the simulation model is fully considered.
And 4, step 4: and (3) excavating implicit information carried by the power plant signal by utilizing a circulating neural network (RNN) in a deep learning algorithm as a supplement to the diagnosis model.
The step 4 further comprises the following steps:
to improve the difficulty of RNN building long-range dependencies, a gating mechanism, i.e., a loop unit (GRU), is introduced. The update gate in the GRU is used to determine which history information is passed backwards, the reset gate is used to determine the degree of discarded history information, and important information in the input sequence is memorized by the network under the gating mechanism, so that the RNN can more easily grasp the long-term dependence between the input sequences.
And 5: and synchronously comparing the MAAP calculation result with the power plant signal, and confirming whether the tracking calculation is consistent with the state of the unit.
The step 5 further comprises the following steps:
if the state tracking consistency is good and no typical event response occurs, then the MAAP is fine-tuned or not adjusted to a small extent according to the current plant signal and the diagnostic result. Otherwise, the MAAP is reinitialized.
And 6: potential intervention measures in a nuclear power plant Serious Accident Management Guide (SAMG) are supported to be introduced into prediction calculation, and positive evaluation, negative evaluation and long-term attention items of the intervention measures are combined with prediction calculation results, so that dynamic evaluation of implementation effects of the intervention measures is realized. And supports simultaneous analysis and comparison of four different intervention protocols.
The step 6 further comprises:
and generating a corresponding prediction report according to the comparative analysis of parameters between different accident processes and the comparative analysis of the accident consequences at different moments and under different intervention schemes to judge the effectiveness of implementation of intervention measures, and providing technical reference for emergency personnel to make a decision list.
The system comprises nuclear power plant operation diagnosis, real-time tracking and rapid prediction, and accords with the design of a third-generation advanced passive nuclear power plant; applying a neural network algorithm to the power plant operation diagnosis; meanwhile, in tracking calculation, real-time simulation is carried out on the operation state of the power plant by using severe accident analysis software MAAP and power plant operation data correction boundary conditions; and in the prediction calculation, predicting the future state of the power plant by using the severe accident analysis software MAAP and combining the current tracking calculation result as input conditions. The technical scheme claimed in the claims comprises complete technical characteristics that the nuclear power plant operation diagnosis, real-time tracking and rapid prediction functions are applied to the design of the advanced passive nuclear power plant suitable for three generations, a neuron network algorithm is applied to the power plant operation diagnosis, the severe accident analysis software MAAP is used for correcting boundary conditions by combining power plant operation data in tracking calculation to simulate the power plant operation state in real time, and the severe accident analysis software MAAP is used for predicting the future state of the power plant by combining the current tracking calculation result as an input condition in the prediction calculation.
Claims (10)
1. An accident online diagnosis tracking and rapid prediction system suitable for a nuclear power plant is characterized in that: the system comprises a nuclear accident diagnosis module, a nuclear accident tracking module and a nuclear accident rapid prediction and intervention measure evaluation module.
2. The system for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants of claim 1, wherein the nuclear accident diagnosis module comprises:
(1) Read and process power plant data, handle a small amount of signals of gathering from the power plant, include: reading power plant rough data, checking whether power plant signals are in a reasonable range, checking the consistency of multiple signal channels of the power plant, completing the combination processing of parameter signal averaging, identifying damaged steam generator loops, combining high-pressure and low-pressure safety injection data, and completing signal unit conversion matched with a bottom layer analysis program;
(2) And (3) overall state identification: identifying the current operating state of the power plant according to the input power plant signals, and judging the severity of the accident, wherein the severity of the accident mainly comprises reactor core submergence, partial reactor core exposure, complete reactor core exposure, reactor core damage, pressure vessel failure and containment failure;
(3) The method comprises the steps of identifying accident types in detail, identifying signs and types of accidents, identifying events occurring in previous and current power plant data sampling intervals, and setting event sequences and boundary conditions of a thermal hydraulic calculation engine by evaluating mass-energy change rates of a steam generator, a main system and a containment and checking whether the containment fails.
3. The system for on-line diagnostic tracking and rapid prediction of accidents at nuclear power plants of claim 1, wherein the nuclear accident tracking module:
(1) Establishing and operating a thermal hydraulic program tracking process, calling and controlling the calculation of a thermal hydraulic calculation engine, and finishing the calculation of an accident state based on input data, a power plant state and an accident category identification result as input;
(2) Adjusting boundary conditions of a thermal hydraulic program, and if the state tracking of the power plant has serious deviation, initializing calculation; if the state of the power plant is tracked well and no typical event response occurs, fine adjustment or no adjustment is carried out on the system parameters in a small range according to the real-time data of the power plant;
(3) And comparing the tracking result with the power plant data, comparing the power plant data with the calculation value of the thermodynamic calculation engine, checking the comparison result to determine whether the tracking module needs to be reinitialized or not, if so, performing assignment of an analysis program according to the current result.
4. The system for on-line diagnostic tracking and rapid prediction of accidents at nuclear power plants of claim 1, wherein the nuclear accident rapid prediction and intervention measure evaluation module:
(1) Acquiring the current unit state calculated by the tracking module as an initial condition of a thermodynamic and hydraulic calculation engine in the prediction module;
(2) By means of the characteristic that a thermodynamic and hydraulic calculation engine in a prediction module can calculate time by time, the future development of the accident is rapidly predicted;
(3) Evaluating implemented intervention measures based on a severe accident management guide rule used by the nuclear power plant;
(4) And the prediction of a plurality of servers is supported, and the comparative evaluation of the accident evolution conditions under different intervention schemes and assumed conditions is realized.
5. The accident online diagnosis tracking and rapid prediction method suitable for the nuclear power plant is characterized by comprising the following steps:
step 1: the nuclear accident diagnosis module is driven by using limited power plant signals as data of a system, identifies signs and types of accidents, identifies events occurring in previous and current power plant data sampling intervals, evaluates the area and position of LOCA, checks whether a containment vessel fails or not and the like as diagnosis results;
and 2, step: the nuclear accident tracking module is used for correcting, calculating and analyzing an event sequence and boundary conditions of an engine in real time according to the power plant signal and the diagnosis result, and carrying out online tracking on the running state of the accident power plant;
and step 3: the nuclear accident rapid prediction and intervention measure evaluation module takes the diagnosis and tracking result as a calculation initial condition, utilizes the MAAP (severe accident analysis software) to predict the state of a future power plant in advance, realizes early warning on a key accident event and supports intervention measure insertion prediction calculation;
and 4, step 4: by utilizing a circulating neural network in a deep learning algorithm, excavating implicit information carried by a power plant signal to supplement a diagnosis model;
and 5: synchronously comparing the MAAP calculation result with the power plant signal, and determining whether the tracking calculation is consistent with the unit state;
step 6: the method supports the introduction of the potential intervention measures into the prediction calculation in the nuclear power plant serious accident management guide rule, combines the positive evaluation, the negative evaluation and the long-term concern items of the intervention measures with the prediction calculation result, realizes the dynamic evaluation of the implementation effect of the intervention measures, and supports the simultaneous analysis and comparison of four different intervention schemes.
6. The method for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants according to claim 5, wherein the step 1 further comprises:
and (4) preprocessing the data of the power plant, reading limited monitorable signals of the power plant, carrying out validity check on the read-in power plant signals, identifying and processing abnormal points, and processing the subsequent analysis and calculation of a data driving system.
7. The method for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants according to claim 5, wherein the step 2 further comprises:
and the nuclear accident tracking module adopts a severe accident analysis software MAAP and builds a physical model of the nuclear power plant.
8. The method for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants according to claim 5, wherein the step 3 further comprises:
the nuclear accident rapid prediction and intervention measure evaluation module also adopts a severe accident analysis software MAAP and builds a physical model of the nuclear power plant;
by means of calculating and analyzing the super real-time performance of the engine, predicting the subsequent development of the accident, and predicting the occurrence time of important events, namely predicting the occurrence time of the important events (such as core melting, pressure vessel lower end socket failure and containment failure);
and aiming at 5 different assumptions of optimal estimation, conservative estimation, optimistic estimation, conservative estimation and optimal estimation, and optimistic estimation, a multi-thread parallel computing mode is adopted.
9. The method for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants according to claim 5, wherein the step 5 further comprises:
if the state tracking consistency is good and no typical event response occurs, performing small-range fine adjustment or no adjustment on the MAAP according to the current power plant signal and the diagnosis result, otherwise, performing reinitialization on the MAAP.
10. The method for on-line diagnosis, tracking and rapid prediction of accidents at nuclear power plants according to claim 5, wherein the step 6 further comprises:
and generating a corresponding prediction report according to the comparison and analysis of parameters among different accident processes and the comparison and analysis of the accident consequences at different moments and under different intervention schemes to judge the effectiveness of the implementation of intervention measures.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211224907.7A CN115798760A (en) | 2022-10-09 | 2022-10-09 | Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211224907.7A CN115798760A (en) | 2022-10-09 | 2022-10-09 | Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115798760A true CN115798760A (en) | 2023-03-14 |
Family
ID=85432630
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211224907.7A Pending CN115798760A (en) | 2022-10-09 | 2022-10-09 | Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115798760A (en) |
-
2022
- 2022-10-09 CN CN202211224907.7A patent/CN115798760A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Coble et al. | Applying the general path model to estimation of remaining useful life | |
| CN110738274A (en) | nuclear power device fault diagnosis method based on data driving | |
| CN112199890B (en) | System-level fault diagnosis method for integrated nuclear power device | |
| JP3068091B1 (en) | Fault diagnosis device | |
| KR102241650B1 (en) | Deep-learning based Reheater Tube Leak detection Method for Thermoelectric power plant And Apparatus Of Thereof | |
| Mathew et al. | Regression kernel for prognostics with support vector machines | |
| CN111354496B (en) | Nuclear power plant accident online diagnosis and state tracking prediction method | |
| CN113657454A (en) | Autoregressive BiGRU-based nuclear power rotating machine state monitoring method | |
| CN114519923A (en) | Intelligent diagnosis and early warning method and system for power plant | |
| CN114925718A (en) | Transformer core grounding current online prediction method and device, storage medium and fault diagnosis method | |
| CN112632845A (en) | Data-based mini-reactor online fault diagnosis method, medium and equipment | |
| Hu et al. | A hybrid model for predicting the degradation trend of hydropower units based on deep learning | |
| CN112016193B (en) | Online prediction method and system for lubrication failure of shield tunneling machine system | |
| CN115600695B (en) | Fault diagnosis method for metering equipment | |
| US11339763B2 (en) | Method for windmill farm monitoring | |
| CN115798760A (en) | Accident online diagnosis tracking and rapid prediction system and method suitable for nuclear power plant | |
| CN116840658A (en) | Fault prediction and diagnosis method for chip firmware | |
| CN114923287B (en) | High-voltage heater fault early warning method, system and medium | |
| CN114320773B (en) | Wind turbine generator system fault early warning method based on power curve analysis and neural network | |
| CN112260211B (en) | Method and device for determining operating state of relay protection device | |
| CN114420333A (en) | Accident diagnosis tracking and rapid prediction system of third-generation passive nuclear power plant | |
| Wang et al. | Complex equipment diagnostic reasoning based on neural network algorithm | |
| CN113837535B (en) | Method for backtracking severe accident process of nuclear power plant | |
| Hines et al. | Application of monitoring and prognostics to small modular reactors | |
| KR20240003469A (en) | Predictive diagnosis method and system of nuclear power plant equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |