CN106874896B - Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system - Google Patents
Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system Download PDFInfo
- Publication number
- CN106874896B CN106874896B CN201710212463.8A CN201710212463A CN106874896B CN 106874896 B CN106874896 B CN 106874896B CN 201710212463 A CN201710212463 A CN 201710212463A CN 106874896 B CN106874896 B CN 106874896B
- Authority
- CN
- China
- Prior art keywords
- time point
- new time
- vector space
- module
- operation parameters
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/192—Recognition using electronic means using simultaneous comparisons or correlations of the image signals with a plurality of references
- G06V30/194—References adjustable by an adaptive method, e.g. learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Testing And Monitoring For Control Systems (AREA)
- Monitoring And Testing Of Nuclear Reactors (AREA)
Abstract
The invention provides an auxiliary learning method for signal feature identification of a nuclear power station loop component loosening diagnosis system, which comprises the following steps of: calculating operating parameters of a plurality of time points; respectively mapping the operation parameters of a plurality of time points to a high-dimensional vector space; adding state labels to a plurality of time points respectively; classifying a plurality of time points in a high-dimensional vector space; inputting and calculating the operation parameters of the new time point, and mapping the operation parameters of the new time point to a high-dimensional vector space; if the vector space distance between the new time point and any time point added with the state label is smaller than a set threshold value, determining that the new time point represents a known event; and if the vector space distance is greater than or equal to the set threshold, giving an alarm to remind a field engineer to check nuclear power working log records of a new time point or to carry out field check on the equipment. The invention also provides an auxiliary learning system. The method and the system can integrate the actual judgment experience of engineers and optimize the alarm parameters.
Description
Technical Field
The invention relates to the field of loose part diagnosis of a nuclear power station, in particular to an auxiliary learning method and system for signal feature identification of a loose part diagnosis system of a nuclear power station primary circuit.
Background
The Loose component diagnosis and vibration Monitoring System (LPMS) for the nuclear power station is called KIR in China, is standard configuration of the current pressurized water reactor nuclear power station, is used for judging the possible Loose of components of a primary circuit System of the nuclear power station and metal components left in the primary circuit during installation, material changing or maintenance.
The loose part can cause material wear if hitting repeatedly, still can cause the fluidic partial obstruction in fuel passage simultaneously, the damage of footing to and the hidden danger that the control rod is stuck, also can cause the accumulation of radioactive substance in the return circuit, and the hidden danger of revealing of return circuit to two.
The existing loose part diagnosis system can analyze the RMS event trigger function and the basic filtering function, can also analyze the trend of background noise history, accurately classify the known time by comparing the background noise, and can automatically ignore the alarm generated by the movement of a control rod; and normal waveforms and impact waveforms can be customized, so that the false alarm rate and the false alarm rate of alarming are reduced. Meanwhile, in order to prevent false alarm, after software alarm, the alarm trigger is not directly sent to the main control room, but sent to the alarm box, and is manually confirmed by a field engineer and then processed.
However, the judgment logic and alarm parameters of the loosening component of the existing systems are preset before the factory shipment. In actual operation, an experienced field engineer can set and adjust alarm parameters according to actual conditions, determine corresponding physical characteristics through parameter change adjustment, and then search for optimized alarm parameter settings according to the physical characteristic changes. Because the experience of an engineer cannot be fused in the conventional system, the loosening component judgment is more accurately carried out and the accuracy of alarm parameters is improved, so that in actual operation, the engineer needs to judge on the basis of the conventional system and the actual experience every time to well make the final loosening component judgment, and the work of the part is accompanied by a large amount of repeated labor, so that the operation efficiency of the nuclear power station is seriously reduced.
Disclosure of Invention
The invention aims to solve the technical problem that the accuracy of an alarm parameter cannot be improved due to the fact that the existing system cannot fuse the experience of an engineer, and provides an auxiliary learning method and an auxiliary learning system for signal feature recognition of a nuclear power station primary loop component loosening diagnosis system, which can fuse the actual judgment experience of the engineer and optimize the alarm parameter.
The technical scheme adopted by the invention for solving the technical problems is as follows: the method for assisting in learning the signal feature identification of the nuclear power plant loop component loosening diagnosis system comprises the following steps:
s1, calculating the operation parameters of a plurality of time points in the operation cycle;
s2, mapping the operation parameters of the time points to a high-dimensional vector space respectively, wherein each time point is represented by a point in the high-dimensional vector space respectively;
s3, respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
s4, classifying a plurality of time points in the high-dimensional vector space;
s5, recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
s6, if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold, determining that the new time point represents a known event, and adding the state label to the new time point; if the vector space distance is greater than or equal to the set threshold, executing step S7;
and S7, giving an alarm, and reminding a field engineer to check the nuclear power work log record of the new time point or perform field check on the equipment.
Preferably, the step S7 further includes the following sub-steps:
s71, if the equipment is judged to normally operate after checking, adding a state label to the new time point, and judging the property of the operating parameter of the new time point recorded in the step S5 to judge the operating parameter as a non-alarm event;
s72, if the operation of the equipment is judged to be abnormal after the check, adding a state tag containing abnormal content to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an alarm event;
and S73, if the operation of the equipment cannot be judged to be abnormal after the check, adding a state label containing an uncertain state to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an alarm event.
Preferably, after the step S7 is performed, the steps S5, S6 and S7 are continuously repeated, and optimization of the correct rate of identification of the loosening member is achieved.
Preferably, in step S1, the operation parameters include frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length;
the operation parameters also comprise median and variance of the operation parameters (namely frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length effective value) of 1 hour, 1 day and 1 week before and after each time point in the operation period, and distance between the operation parameters of each time point in the operation period and the median of the operation parameters of 1 hour, 1 day and 1 week before and after the operation period.
Preferably, the status tags include yes/no part release event, yes/no device abnormality, hot test, start-up, shut-down of cooling pump, and various temperature and pressure steps of temperature and pressure increase and decrease in steps S3 and S6.
The invention also provides an auxiliary learning system for signal feature recognition of a nuclear power station loop component loosening diagnosis system, which comprises the following modules:
the calculation module is used for calculating the operation parameters of a plurality of time points in the operation period;
a mapping module, configured to map the operating parameters of the multiple time points to a high-dimensional vector space, where each time point is represented by a point in the high-dimensional vector space;
the state label module is used for respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
a classification module for classifying the plurality of time points in the high-dimensional vector space;
the recording module is used for recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
the judging module is used for determining that the new time point represents a known event and adding a state label to the new time point if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold; if the vector space distance is larger than or equal to a set threshold value, executing an alarm module;
and the alarm module is used for giving an alarm and then reminding a field engineer to check the nuclear power working log record of the new time point or carry out field check on the equipment.
Preferably, the alarm module comprises the following sub-modules:
the first judgment alarm module is used for adding a state label to the new time point and judging the property of the operation parameter of the new time point recorded in the recording module if the equipment is judged to operate normally after the examination, and judging the operation parameter as a non-alarm event;
the second judgment and alarm module is used for adding a state label containing abnormal content to the new time point if the equipment is judged to be abnormal after the equipment is checked, judging the property of the operation parameter of the new time point input into the module and judging the operation parameter as an alarm event;
and the third judgment alarm module is used for adding a state label containing an uncertain state to the new time point if the operation of the equipment cannot be judged to be abnormal after the check, judging the property of the operation parameter of the new time point recorded in the recording module and judging the operation parameter as an uncertain event.
Preferably, after the alarm module is executed, the logging module, the judging module and the alarm module are continuously and repeatedly executed, so that the identification accuracy of the loosening component is optimized.
Preferably, in the calculation module, the operation parameters include frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length;
the operation parameters also comprise median and variance of the operation parameters (namely frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length effective value) of 1 hour, 1 day and 1 week before and after each time point in the operation period, and distance between the operation parameters of each time point in the operation period and the median of the operation parameters of 1 hour, 1 day and 1 week before and after the operation period.
Preferably, in the status label module and the judgment module, the status label includes yes/no component release event, yes/no equipment abnormality, hot test, start-up, shut-down of the cooling pump, and temperature and pressure steps for temperature and pressure increase and decrease.
The auxiliary learning method and the auxiliary learning system have the advantages that in actual operation, an experienced field engineer can set and adjust the alarm parameters according to actual conditions, corresponding physical characteristics are determined by adjusting parameter changes, and then the alarm parameter setting is adjusted and optimized according to the physical characteristic changes. The method and the system provided by the invention integrate the actual judgment experience of engineers, reduce a large amount of unnecessary repeated labor, optimize alarm parameters and greatly improve the diagnosis efficiency of loose parts.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a schematic flow chart of an auxiliary learning method for signal feature recognition of a loose diagnosis system of a loop component of a nuclear power plant according to the present invention;
FIG. 2 is a flow chart illustrating a sub-step of step S7 in the method for learning the signal feature identification of the loose component diagnosis system in a nuclear power plant according to the present invention;
fig. 3 is a schematic diagram of mapping of operation parameters to a high-dimensional vector space in an embodiment of an auxiliary learning method for signal feature identification of a loose diagnosis system of a primary loop component of a nuclear power plant according to the present invention.
Detailed Description
For a more clear understanding of the technical features, objects and effects of the present invention, embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
FIG. 1 is a schematic flow chart of an auxiliary learning method for signal feature recognition of a nuclear power plant primary loop component loosening diagnosis system. The auxiliary learning method comprises the following steps:
s1, calculating the operation parameters of a plurality of time points in the operation cycle;
s2, mapping the operation parameters of the time points to a high-dimensional vector space respectively, wherein each time point is represented by a point in the high-dimensional vector space respectively;
s3, respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
s4, classifying a plurality of time points in the high-dimensional vector space;
s5, recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
s6, if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold, determining that the new time point represents a known event, and adding the state label to the new time point; if the vector space distance is greater than or equal to the set threshold, executing step S7;
and S7, giving an alarm, and reminding a field engineer to check the nuclear power work log record of the new time point or perform field check on the equipment.
The operation cycle refers to a cycle of the whole process of refueling, temperature and pressure rising, temperature and pressure reduction and overhaul of the nuclear power plant. The plurality of time points are time points which are set in advance according to needs through time intervals and obtained through continuous sampling, the time intervals among the time points are the same, the length of the time intervals can be freely set according to needs, and the time intervals can be 30 seconds, one minute, five minutes, half hour and the like.
In step S1, the operating parameters include frequency spectrum, signal harmonics, shock signal frequency, shock signal decay time, sensor signal delay, RMS amplitude, and RMS ratio.
The operation parameters also comprise median and variance of the operation parameters (namely frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length effective value) of 1 hour, 1 day and 1 week before and after each time point in the operation period, and distance between the operation parameters of each time point in the operation period and the median of the operation parameters of 1 hour, 1 day and 1 week before and after the operation period. For example, if the signal spectrum at a time point is less distant from the median of the previous and subsequent operating parameters of 1 hour, 1 day and 1 week, and belongs to a certain specific working condition, the condition that the component is loosened can be eliminated; if the impact signal frequency and the impact signal attenuation time at the time point are close to the median distances of the impact signal frequency and the impact signal attenuation time at the time points of 1 hour, 1 day and 1 week before and after, respectively, and are within the interval range of the set threshold, the probability of the component loosening condition is high.
In steps S3 and S6, the status labels include yes/no part release events, yes/no device exceptions, warm-up, start-up, shut-down cooling pump, and various temperature and pressure steps for warming up and cooling down.
In step S2, if the mapped high-dimensional vector space does not include a time axis, each time point is represented by a point in the high-dimensional vector space, and if the mapped high-dimensional vector space includes a time axis, each time point has a corresponding time on the time axis, and the rest of the operating parameters are represented by points in the high-dimensional vector space.
Fig. 2 is a flow chart illustrating a sub-step of step S7 in the method for assisting learning of signal feature identification of a nuclear power plant primary loop component loosening diagnosis system according to the present invention. Step S7 further includes the following sub-steps:
s71, if the equipment is judged to normally operate after checking, adding a state label to the new time point, and judging the property of the operating parameter of the new time point recorded in the step S5 to judge the operating parameter as a non-alarm event;
s72, if the operation of the equipment is judged to be abnormal after the check, adding a state tag containing abnormal content to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an alarm event;
and S73, if the operation of the equipment cannot be judged to be abnormal after the check, adding a state label containing an uncertain state to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an uncertain event.
After the step S7 is executed, the steps S5, S6 and S7 are continuously repeated, and optimization of the correct recognition rate of the loosening member is achieved. In the existing system, the KIR loose part diagnosis system of the nuclear power plant only judges and alarms the RMS amplitude of the time-domain waveform. The alarm parameters are set and then cannot be changed, so that the identification rate of the diagnosis system to loose parts in a loop is fixed, and the identification accuracy cannot be improved along with the increase of the operation time of the system and the accumulation of historical data. According to the method, by adding the marking and judgment of a field engineer on the alarm time signal, historical data are accumulated, and the accuracy of the identification rate of loose parts or other abnormal events is improved based on the increasing historical data.
The method has another important significance in that after the marking and judgment of high-grade engineers with abundant experience on the historical data are added into the method and the system, the probability of errors in alarming and judging can be greatly reduced when the engineers with less experience face again in the future, and the safety of nuclear power operation is critically guaranteed.
In applying the method of the present invention, the larger the base number of the plurality of time points in step S1, the more the number of iterations of steps S5-S7 are repeated, the more accurate the identification of whether the component loosening phenomenon or other abnormality occurs at the next new time point.
The plurality of time points in step S1 includes the entire operation cycle.
An example of a listing of the plurality of time points in step S1 is shown in the following table:
TABLE 1
Step S2 of the present invention maps the operating parameters for each of the time points in table 1 to a high-dimensional vector space, each time point being represented by a point in the vector space. The time points are classified by the distance in the vector space in step S4. If the spatial distance between one time point and the time point with the closest vector spatial distance is less than a set threshold value, the two points are classified into one class; if the spatial distance between a time point and the time point with the closest vector spatial distance is greater than or equal to a set threshold value, the two points are determined to be in different classes.
Fig. 3 is a schematic diagram of mapping of operation parameters to a high-dimensional vector space in an embodiment of an auxiliary learning method for signal feature identification of a loose diagnosis system of a primary loop component of a nuclear power plant according to the present invention. In the embodiment of fig. 3, only three operating parameters for each time point are used for mapping, so that a three-dimensional vector space is mapped. The first operating parameter is the ratio of RMS magnitude, the second operating parameter is the frequency of the shocksignal, and the third operating parameter is the frequency spectrum of the shocksignal. After the method in step S4 is applied to the time points in fig. 3 for classification, it can be seen that the time points in the figure can be classified into three categories according to the spatial distance.
The invention also provides an auxiliary learning system for signal feature recognition of a nuclear power station loop component loosening diagnosis system, which comprises the following modules:
the calculation module is used for calculating the operation parameters of a plurality of time points in the operation period;
a mapping module, configured to map the operating parameters of the multiple time points to a high-dimensional vector space, where each time point is represented by a point in the high-dimensional vector space;
the state label module is used for respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
a classification module for classifying the plurality of time points in the high-dimensional vector space;
the recording module is used for recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
the judging module is used for determining that the new time point represents a known event and adding a state label to the new time point if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold; if the vector space distance is larger than or equal to a set threshold value, executing an alarm module;
and the alarm module is used for giving an alarm and then reminding a field engineer to check the nuclear power working log record of the new time point or carry out field check on the equipment.
The alarm module comprises the following sub-modules:
the first judgment alarm module is used for adding a state label to the new time point and judging the property of the operation parameter of the new time point recorded in the recording module if the equipment is judged to operate normally after the examination, and judging the operation parameter as a non-alarm event;
the second judgment and alarm module is used for adding a state label containing abnormal content to the new time point if the equipment is judged to be abnormal after the equipment is checked, judging the property of the operation parameter of the new time point input into the module and judging the operation parameter as an alarm event;
and the third judgment alarm module is used for adding a state label containing an uncertain state to the new time point if the operation of the equipment cannot be judged to be abnormal after the check, judging the property of the operation parameter of the new time point recorded in the recording module and judging the operation parameter as an uncertain event.
After the alarm module is executed, the input module, the judgment module and the alarm module are continuously and repeatedly executed, and the correct identification rate of the loosening component is optimized.
In the calculation module, the operation parameters include frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time and sensor signal delay.
In the status label module and the judgment module, the status label comprises a yes/no component loosening event, a yes/no equipment abnormity, a hot test, starting, stopping a cooling pump and temperature and pressure steps for heating, cooling and pressing.
While the present invention has been described with reference to the embodiments shown in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (6)
1. An auxiliary learning method for signal feature recognition of a nuclear power station primary loop component loosening diagnosis system is characterized by comprising the following steps:
s1, calculating the operation parameters of a plurality of time points in the operation cycle;
s2, mapping the operation parameters of the time points to a high-dimensional vector space respectively, wherein each time point is represented by a point in the high-dimensional vector space respectively;
s3, respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
s4, classifying a plurality of time points in the high-dimensional vector space;
s5, recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
s6, if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold, determining that the new time point represents a known event, and adding the state label to the new time point; if the vector space distance is greater than or equal to the set threshold, executing step S7;
s7, giving an alarm, and reminding a field engineer to check the nuclear power working log record of the new time point or to perform field check on equipment;
in step S1, the operation parameters include frequency spectrum, signal harmonic, impulse signal frequency, impulse signal attenuation time, sensor signal delay, RMS amplitude, and ratio of RMS length to RMS value; the operation parameters also comprise the median and the variance of the operation parameters of 1 hour, 1 day and 1 week before and after each time point in the operation period, and the distance between the operation parameters of each time point in the operation period and the median of the operation parameters of 1 hour, 1 day and 1 week before and after the operation period;
in steps S3 and S6, the status labels include yes/no part release events, yes/no device exceptions, warm-up, start-up, shut-down cooling pump, and various temperature and pressure steps for warming up and cooling down.
2. The method for assisting in learning the signal feature identification of the nuclear power plant primary loop component loosening diagnosis system as claimed in claim 1, wherein the step S7 further comprises the following sub-steps:
s71, if the equipment is judged to normally operate after checking, adding a state label to the new time point, and judging the property of the operating parameter of the new time point recorded in the step S5 to judge the operating parameter as a non-alarm event;
s72, if the operation of the equipment is judged to be abnormal after the check, adding a state tag containing abnormal content to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an alarm event;
and S73, if the operation of the equipment cannot be judged to be abnormal after the check, adding a state label containing an uncertain state to the new time point, and judging the property of the operation parameter of the new time point recorded in the step S5 to be an uncertain event.
3. The auxiliary learning method for signal feature identification of a nuclear power plant loop component loosening diagnosis system as claimed in claim 2, wherein after the step S7 is executed, the steps S5, S6 and S7 are continuously repeated to optimize the correct rate of loosening component identification.
4. An auxiliary learning system for signal feature recognition of a nuclear power station primary loop component loosening diagnosis system is characterized by comprising the following modules:
the calculation module is used for calculating the operation parameters of a plurality of time points in the operation period;
a mapping module, configured to map the operating parameters of the multiple time points to a high-dimensional vector space, where each time point is represented by a point in the high-dimensional vector space;
the state label module is used for respectively adding state labels to the time points according to nuclear power working log records and operation parameters;
a classification module for classifying the plurality of time points in the high-dimensional vector space;
the recording module is used for recording and calculating the operation parameters of a new time point, and mapping the operation parameters of the new time point to the high-dimensional vector space, wherein the new time point is represented by points in the high-dimensional vector space;
the judging module is used for determining that the new time point represents a known event and adding a state label to the new time point if the vector space distance between the new time point and any of the plurality of time points added with the state label is smaller than a set threshold; if the vector space distance is larger than or equal to a set threshold value, executing an alarm module;
the alarm module is used for giving an alarm and then reminding a field engineer to check the nuclear power working log record of the new time point or carry out field check on the equipment;
in the calculation module, the operation parameters comprise frequency spectrum, signal harmonic, impact signal frequency, impact signal attenuation time, sensor signal delay, RMS amplitude and ratio of RMS length to RMS length; the operation parameters also comprise the median and the variance of the operation parameters of 1 hour, 1 day and 1 week before and after each time point in the operation period, and the distance between the operation parameters of each time point in the operation period and the median of the operation parameters of 1 hour, 1 day and 1 week before and after the operation period;
in the status label module and the judgment module, the status label comprises a yes/no component loosening event, a yes/no equipment abnormity, a hot test, starting, stopping a cooling pump and temperature and pressure steps for heating, cooling and pressing.
5. The system of claim 4, wherein the alarm module comprises the following sub-modules:
the first judgment alarm module is used for adding a state label to the new time point and judging the property of the operation parameter of the new time point recorded in the recording module if the equipment is judged to operate normally after the examination, and judging the operation parameter as a non-alarm event;
the second judgment and alarm module is used for adding a state label containing abnormal content to the new time point if the equipment is judged to be abnormal after the equipment is checked, judging the property of the operation parameter of the new time point input into the module and judging the operation parameter as an alarm event;
and the third judgment alarm module is used for adding a state label containing an uncertain state to the new time point if the operation of the equipment cannot be judged to be abnormal after the check, judging the property of the operation parameter of the new time point recorded in the recording module and judging the operation parameter as an uncertain event.
6. The system for assisting in learning signal feature identification of a nuclear power plant loop component loosening diagnosis system as claimed in claim 5, wherein the logging module, the judging module and the alarming module are continuously and repeatedly executed after the alarming module is executed, so as to optimize the accuracy of loosening component identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710212463.8A CN106874896B (en) | 2017-03-31 | 2017-03-31 | Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710212463.8A CN106874896B (en) | 2017-03-31 | 2017-03-31 | Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106874896A CN106874896A (en) | 2017-06-20 |
| CN106874896B true CN106874896B (en) | 2021-03-02 |
Family
ID=59160107
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710212463.8A Active CN106874896B (en) | 2017-03-31 | 2017-03-31 | Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106874896B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107544337B (en) * | 2017-09-19 | 2020-10-30 | 中国核动力研究设计院 | Method for intelligently classifying triggering data of loose part monitoring system |
| CN110532836B (en) * | 2018-05-25 | 2022-03-22 | 中广核工程有限公司 | Nuclear power station signal identification method and device, computer equipment and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101487865A (en) * | 2008-01-16 | 2009-07-22 | 西门子公司 | Fault diagnosis method and fault diagnosis system |
| CN101685679A (en) * | 2008-09-27 | 2010-03-31 | 中国核动力研究设计院 | System for monitoring loosening part of nuclear reactor and coolant system |
| CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
| CN101750197A (en) * | 2008-12-11 | 2010-06-23 | 中国核动力研究设计院 | Loose part diagnosis method for nuclear power station |
| CN103137224A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part quality estimation method based on wavelet energy spectrum |
| CN204886920U (en) * | 2015-08-22 | 2015-12-16 | 上海核工程研究设计院 | Not hard up part of nuclear power station signal conditioning circuit for monitoring system |
| CN105467975A (en) * | 2015-12-29 | 2016-04-06 | 山东鲁能软件技术有限公司 | Equipment fault diagnosis method |
| CN105931685A (en) * | 2016-06-14 | 2016-09-07 | 中广核工程有限公司 | Nuclear power station primary circuit foreign matter monitoring system and nuclear power station primary circuit foreign matter monitoring method |
| CN105955069A (en) * | 2016-06-12 | 2016-09-21 | 哈尔滨工程大学 | On-line-simulated-based nuclear power plant system level state monitoring method |
-
2017
- 2017-03-31 CN CN201710212463.8A patent/CN106874896B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101487865A (en) * | 2008-01-16 | 2009-07-22 | 西门子公司 | Fault diagnosis method and fault diagnosis system |
| CN101685679A (en) * | 2008-09-27 | 2010-03-31 | 中国核动力研究设计院 | System for monitoring loosening part of nuclear reactor and coolant system |
| CN101750197A (en) * | 2008-12-11 | 2010-06-23 | 中国核动力研究设计院 | Loose part diagnosis method for nuclear power station |
| CN101718581A (en) * | 2009-11-13 | 2010-06-02 | 浙江大学 | Alarming method of nuclear power station loose part based on support vector machine |
| CN103137224A (en) * | 2013-01-25 | 2013-06-05 | 杭州电子科技大学 | Nuclear power station loose part quality estimation method based on wavelet energy spectrum |
| CN204886920U (en) * | 2015-08-22 | 2015-12-16 | 上海核工程研究设计院 | Not hard up part of nuclear power station signal conditioning circuit for monitoring system |
| CN105467975A (en) * | 2015-12-29 | 2016-04-06 | 山东鲁能软件技术有限公司 | Equipment fault diagnosis method |
| CN105955069A (en) * | 2016-06-12 | 2016-09-21 | 哈尔滨工程大学 | On-line-simulated-based nuclear power plant system level state monitoring method |
| CN105931685A (en) * | 2016-06-14 | 2016-09-07 | 中广核工程有限公司 | Nuclear power station primary circuit foreign matter monitoring system and nuclear power station primary circuit foreign matter monitoring method |
Non-Patent Citations (4)
| Title |
|---|
| "基于改进归纳式监控算法的液体火箭发动机实时故障检测";高克寒等;《航空动力学报》;20161031;第31卷(第10期);第2556页第1.2-1.3节 * |
| "基于虚拟仪器的核电站松脱部件监测系统";盘世标等;《核动力工程》;20081231;第29卷(第6期);第125-127页 * |
| "松脱部件检测系统在核电站的应用研究";李坤;《核电子学与探测技术》;20150731;第35卷(第1期);第712页第1.1节 * |
| "某化工厂环氧化反应釜爆炸事故原因分析对策措施";张晓民;《化学工程与装备》;20140930(第9期);第222-224页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106874896A (en) | 2017-06-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109524139B (en) | Real-time equipment performance monitoring method based on equipment working condition change | |
| US8988238B2 (en) | Change detection system using frequency analysis and method | |
| US11531688B2 (en) | Time-series data processing device, time-series data processing system, and time-series data processing method | |
| US8248228B2 (en) | Method and device for optimizing the alarm configuration | |
| CN109186813A (en) | A kind of temperature sensor self-checking unit and method | |
| EP3055747A1 (en) | Correlation and annotation of time series data sequences to extracted or existing discrete data | |
| US20150006972A1 (en) | Method for Detecting Anomalies in a Time Series Data with Trajectory and Stochastic Components | |
| CN116593811B (en) | Integrated frequency converter running state monitoring system and monitoring method | |
| US20150241304A1 (en) | Method for the computer-assisted monitoring of the operation of a technical system, particularly of an electrical energy-generating installation | |
| CN104075751A (en) | Internet data center temperature and humidity early warning method and device | |
| CN106874896B (en) | Auxiliary learning method and system for signal feature identification of nuclear power station primary loop component loosening diagnosis system | |
| CN114781476B (en) | System and method for analyzing faults of measuring equipment | |
| CN103675355A (en) | Anemometer monitoring method and system | |
| CN112632845B (en) | Data-based mini-reactor online fault diagnosis method, medium and equipment | |
| CN103197663A (en) | Method and system of failure prediction | |
| CN111596643A (en) | Visual dynamic energy consumption diagnosis, analysis and pre-control system based on big data | |
| CN110287827A (en) | A kind of bridge strain data outliers recognition methods based on data correlation | |
| EP3404589A1 (en) | Computer system and method for improved monitoring of the technical state of industrial systems | |
| CN105425739A (en) | System for predicting abnormality occurrence using PLC log data | |
| CN110991074A (en) | Method for judging validity of measurement data of displacement sensor | |
| KR101917477B1 (en) | Pre-sensing apparatus for abnormal of coiling equipment | |
| CN103675357A (en) | Anemometer fault warning method and system | |
| CN117423225A (en) | Disaster remote sensing early warning system based on high-speed railway operation | |
| CN117279348A (en) | Strong electromagnetic pulse protection system with monitoring function | |
| CN106650154B (en) | Nuclear power conventional island equipment reliability monitoring method based on thermal performance index |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |