CN101067976B - Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method - Google Patents
Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method Download PDFInfo
- Publication number
- CN101067976B CN101067976B CN200610152151A CN200610152151A CN101067976B CN 101067976 B CN101067976 B CN 101067976B CN 200610152151 A CN200610152151 A CN 200610152151A CN 200610152151 A CN200610152151 A CN 200610152151A CN 101067976 B CN101067976 B CN 101067976B
- Authority
- CN
- China
- Prior art keywords
- point
- maximal
- minimal
- enter
- certain
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E30/00—Energy generation of nuclear origin
- Y02E30/30—Nuclear fission reactors
Landscapes
- Testing And Monitoring For Control Systems (AREA)
- Plasma Technology (AREA)
Abstract
This invention provides an on-line monitor and fault diagnosis method for PWR nuclear power plant including the following steps: 1, data collection, 2, digit filtration and pre-process, 3, acquiring extremum points to carry out polygon process to the waveforms, 4, merging the extremum points and eliminating single peaks, 5, identifying character points of the waveforms in terms of upgrading, transmission and keeping to check current and waving amplitude, 6, judging step sequence and operation time and diagnoses fault, 7, comparing the consistency of four rod-beam operations of a sub-set to estimate them according to deviations and finish single-step analysis of the sub-set waveform, 8, collecting and displaying waveforms, storing effective data and storing action recordings.
Description
Technical field
The present invention relates to a kind of inline diagnosis and method for diagnosing faults of on-line monitoring method, particularly a kind of pressurized water reactor nuclear power station control bar driving mechanism.
Background technology
Control rod drive mechanisms (CRD) (hereinafter to be referred as CRDM:Control Rod Drive Mechanism) is unique movable member in the nuclear reactor pressure container, and its fault or performance degradation may cause card rod, sliding excellent accident, have a strong impact on the unit availability.Qinshan second phase 1, No. 2 unit rods control equipment is provided with monitoring circuit at driving mechanism coil current size and ripple.The current failure diagnosis is realized by hardware, do not combine with driving mechanism action command sequential, an excessive alarming value of electric current and the too small alarming value of electric current can only be set, can not monitor that whether big electric current is less than definite value, whether little electric current, zero current do not monitor transient process, operating point, current waveform vibration situation greater than definite value.This function for monitoring a egregious cases at Primary Component complete failure, oscillating waveform, size of current fault that device performance degeneration that most probable is taken place or the drift of crucial setting valve cause do not monitor, cause the situation of operating point fault or forfeiture operating point not monitor to the driving mechanism mechanical fault, can only wait these faults constantly to add up, after making the deviation that generation is bigger between control rod physical location and the given position, come out by the rod position measurement Device Diagnostic.Before the rod position measurement monitoring of equipment arrived, these faults just were damaged driving mechanism, driving shaft probably, perhaps caused cluster to drop and then caused the shutdown incident, caused bigger influence for the nuclear power generating sets safety and economic operation.
In addition, " Nuclear Power Engineering " September in 2002 the 12 national reactor structure mechanics meeting paper special edition 241-243 page or leaf, " control rod electromagnet driving mechanism principle experiment monitoring system " literary composition that the Zhang Jige of nuclear energy technology design and research institute of Tsing-Hua University etc. writes, introduced a kind of brand-new control rod drive mechanisms (CRD) scheme that nuclear energy technology design and research institute of Tsing-Hua University develops alone---employed control and data acquisition equipment in the moving coil solenoid actuated mechanism principle model machine performance test, adopt light to walk slowly like a woman the actual displacement that chi measures control rod of this principle prototype experiment monitoring system, adopt the control of Programmable Logic Controller (the being called for short PLC) platform that experimentizes, adopt the experimentize collection of data of computing machine, communicate by the RS232 serial ports between computing machine and the PLC.But it just sends steering order, Displacement Measurement checking driving mechanism function.
Summary of the invention
The object of the present invention is to provide a kind of inline diagnosis and method for diagnosing faults of having realized on-line fault diagnosis, having guaranteed in time to find the pressurized water reactor nuclear power station control bar driving mechanism of also accurate fault location.
Realize the technical scheme of the object of the invention: a kind of inline diagnosis of pressurized water reactor nuclear power station control bar driving mechanism and method for diagnosing faults may further comprise the steps:
(1) data acquisition;
(2) digital filtering;
(1) establishing the c point is pending point, and a, b are continuous 2 points in its front, and d, e, f are continuous 3 points in its back, judges according to the magnitude relationship of some before and after c and its whether it is unusual;
If min{|c-b|, | c-d|}＞2 * max{|a-b|, | b-d|, | d-e|} then claims c to depart from a, b, d, e, three kinds of situations of calculating value distribution: relatively | a-b|, | b-d|, the little usefulness of the latter (b+d)/2 with b+ (b-a)/2, is used b+ (b-a) when rising as the former small form during decline;
(2) putting preceding two point prediction values is preceding 6 averages;
(3) since the 3rd data, up to third from the bottom point, following processing is carried out in circulation:
Judge whether c ≈ 0 or be lasting rising or intermediate point for continuing to descend? be, then do not deal with; , then do not judge whether to be one of following situation:
Be then to get b, d mean value is the c predicted value;
1. fall greatly and fall: the peak that occurs in the decline process;
2. fall for a short time and fall: the ebb that occurs in the decline process;
3. rise and rise higher: occur the peak during rising;
4. rise little liter: occur ebb during rising;
5. fall and rise higher: the peak appears in the commentaries on classics rising portion that descends;
6. rise and fall for a short time: rising to change ebb occurs when descending;
Not, judge whether that then c departs from a, b, d, f? be, then according to a, b, d, f prediction c value; Not, judge whether that then c departs from a, b, e, f?
Be, then according to a, b, e, f prediction c value; Not, judge whether that then c departs from a, b, d, e?
Be, then according to a, b, d, e prediction c value; Not, judge whether that then c is for going up the summit unusually? c is that last summit and c-b＞b-a think that then c goes up the summit unusually;
Be that then the c value changes b+ (b-a)/2 into; Not, judge whether that then c is for descending the summit unusually? c is that following summit and b-c＞a-b think that then c descends the summit unusually;
Be that then the c value changes 2b-a into; Not, then the c value is constant;
(4) 2 of postpones are the predicted value of 4 of its fronts;
(3) ask for extreme point, waveform is carried out the broken line processing
Local maximum minimal point in the search one dimension scalar array, currency is greater than last value or approximate last value and be in local ascent stage, confirms minimal point when minimal point is unconfirmed, upgrade unique point, short-term trend changes rising into, and minimal point changes into confirms that maximal point changes into unconfirmed; Otherwise then currency upgrades unique point less than last value or approximate last value and be in and local confirm maximal point when the decline stage, maximal point was unconfirmed, and short-term trend changes decline into, and maximal point changes into to be confirmed, it is unconfirmed that minimal point changes into;
(4) extreme point merges and unimodal elimination processing
(1) if certain minimal point, is then deleted next maximal point, next minimal point less than next minimal point;
(2) if certain minimal point 〉=next minimal point, then deletion itself and next maximal point;
(3) if certain maximal point less than next maximal point, then deletion itself and next minimal point;
(4) if certain maximal point 〉=next maximal point is then deleted next minimal point, maximal point;
(5) divide lifting, transmission, three kinds of situations identification of maintenance waveform character points, check size of current and fluctuating range
Adopt the finite state machine pattern to promoting the method that coil current waveform character point is discerned, may further comprise the steps:
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 1A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-40; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 1A, enter (9) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-40;
(3) rising 0-40 when certain minimal point≤0, enters (2) zero current platform; When the average of three maximal points of certain maximal point and back differ＜during 1A, enter (4) adhesive platform;
(4) adhesive platform, three the maximal point averages in back enter (5) big electric current platform than more than three the big 0.5A of maximal point average in front; When behind adhesive platform 100ms, not finding operating point yet, enter (5) big electric current platform; Certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform;
(5) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (6) decline 40-16;
(6) decline 40-16, half of certain minimal point-back minimal point≤back increasing degree enters (7) little electric current platform;
(7) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (8) decline 16-0;
(8) decline 16-0, certain minimal point or thereafter maximal point be about 0, enter (9) and stop;
(9) stop;
Adopt the finite state machine pattern to the method that transfer line loop current ripple carries out unique point identification, may further comprise the steps:
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 0.5A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-8; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 0.5A, enter (10) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-8;
(3) rising 0-8 when certain minimal point≤0, enters (2) zero current platform; When certain maximal point＞last maximal point, simultaneously＞next maximal point-0.1A, enter the preceding decline of (4) adhesive;
(4) descend before the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; When behind adhesive platform 160ms, not finding operating point yet, rise after entering (5) adhesive; Several minimal points before and after a certain minimal point is significantly less than rise after entering (5) adhesive;
(5) rise after the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; A certain maximal point is greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (6) big electric current platform;
(6) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (7) decline 8-4;
(7) decline 8-4, certain minimal point＜1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (10) and stop; , certain minimal point＞1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (8) little electric current platform;
(8) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (9) decline 4-0;
(9) decline 4-0, certain minimal point or thereafter maximal point be about 0, enter (10) and stop;
(10) stop;
Adopt the finite state machine pattern to the method that the holding coil current waveform carries out unique point identification, may further comprise the steps:
(1) original state, the relative wave amplitude of certain minimal point are between 0.2-0.5, and be very little with back minimal point relative deviation, enters (2) little electric current and keep; Two maximal points in certain minimal point back significantly rise and current point＞0.8 times increasing degree continuously, enter (3) rising 4-8; Certain minimal point is about 0, enters (6) zero current platform; Two maximal points in certain minimal point back significantly rise and current point≤0.8 times increasing degree continuously, enter (7) rising 0-8; Two the minimal point＞6A in front, current point＜5A, and back two minimal points＞3.5A enter (12) and stop;
(2) little electric current keeps, and two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 4-8; Last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(3) rising 4-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (4) big electric current and keep; A certain minimal point is about 0, enters (6) zero current platform;
(4) big electric current keeps, and the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enters (5) decline 8-0;
(5) decline 8-0, last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(6) zero current platform, two maximal points in certain minimal point back significantly rise continuously, enter (7) rising 0-8;
(7) rising 0-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, descends before entering (8) adhesive;
(8) descend before the adhesive, behind adhesive platform 100ms, do not find operating point yet, enter (10) big electric current platform; Several minimal points before and after a certain minimal point is significantly less than rise after entering (9) adhesive;
(9) rise after the adhesive, differ half, enter (10) big electric current platform less than the range of decrease with three maximal points in back;
(10) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (11) decline 8-4;
(11) decline 8-4,0.5 times of minimal point one back minimal point≤back increasing degree enters (12) and stops;
(12) stop;
(6) differentiate and fault diagnosis step preface, actuation time, lifting coil current waveform, transfer line loop current waveform, holding coil current waveform are carried out analyzing and processing, form the operating point array, judge whether to be normal step preface, finish single CRDM step preface inspection;
(7) consistance to 4 cluster actions of a son group compares, and carries out fault diagnosis according to the deviation size, finishes the single step analysis of son group waveform;
(8) gather waveform demonstration, valid data storage, action record storage.
Effect of the present invention: (1) adopts Wave data broken line disposal route.Be similar to broken line by asking for the local extremum current waveform of naming a person for a particular job, significantly reduce the core feature that needs the data volume of handling and extract waveform, merge to handle by digital filtering and broken line end points and eliminate pseudo-data.(2) adopt the finite state machine algorithm to extract the waveform character point.Normal waveform is divided into several states, set specific state transition condition, handle by the effective status machine through the collection waveform after the broken line processing, judge according to final state and the operating point data that recognize whether the step preface is complete, whether wave amplitude, size of current be normal.(3) adopt the virtual instrument technology systemic-function.This method has realized on-line fault diagnosis, has guaranteed in time to find and accurate fault location; and fault, waveform unusual or that comprise operating point preserved, show; can long-term accumulation CRDM data actuation time, carry out trend analysis, be beneficial to and find the CRDM fault symptom of a trend as early as possible.
Description of drawings
Fig. 1 is the process flow diagram of pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and method for diagnosing faults.
Fig. 2 is digital filtering and pretreated process flow diagram;
Fig. 3 is the synoptic diagram of singular point filtering algorithm;
Fig. 4 is the design sketch of singular point filtering algorithm;
Fig. 5 is waveform broken lineization, the design sketch that merges processing;
Fig. 6 is for promoting the process flow diagram of coil current waveform character identification state machine;
Fig. 7 is the process flow diagram of transfer line loop current waveform character identification state machine;
Fig. 8 is the process flow diagram of holding coil Current Waveform Characteristics identification state machine.
Embodiment
Qinshan second stage reaction heap control rod has 33 bundles.8 shutdown clusters are regulated cluster for 25, are used for conditioned reaction heap power.8 shutdown clusters are weaved into a rod group---and Sa, Sa are divided into Sa1 and two son groups of Sa2 again, and each son group comprises 4 clusters.Regulate the cluster volume for 25 and be A, B, four rod groups of C, D, A, B, three rods of C are organized and respectively are divided into 1,2 two son group; A1, A2, B1, B2, C1, D respectively contain 4 clusters, and C2 only contains 1 cluster.33 of correspondingly total control rod drive mechanisms (CRD), each son group is by power cabinet control, totally 9 power cabinets.
Every driving mechanism has 3 coils, needs to measure 3 current signals, if add the sound vibration signal of decision mechanism action, then each driving mechanism has 4 signal demand collections.It is 132 that 33 driving mechanism action monitorings need the signal of collection, convenience for data processing, be convenient to simultaneously four mechanism action consistance of a son group are compared, 16 signals to four driving mechanisms of a son group carry out acquisition process simultaneously, to having only the C2 group of a mechanism, during software processes its all signals one are divided into four, so in fact the number of signals of Chu Liing is 144.
The CRDM coil current is that the AC power of 50Hz forms through three phase half wave rectification, and ripple frequency is 150Hz.Judge failure conditions such as dephasing, current ripples be too big for correct, sampling period less than 1 millisecond better, get 0.5 millisecond, corresponding sample frequency is 2000Hz, it is 16 * 2000 that each son is organized the data volume in 1 second, double-precision floating points storage in full according to this, 8 bytes of each data, need about disk space 250KB 9 about 1.8MB of son group per second peak-data altogether.Stroke of all son group operations (450 step) data volume is 9 * 450 * 200KB=810MB.In fact the unit normal operation period has only the action of D group control rod, moves every day about 20 steps, and the data volume that need write down seldom.If a son group waveform is unusual, need continuously record data, then a day data amount is 250KB * 60 * 60 * 24=21.6G.
According to above data processing requirement, select the PXI of American National instrument company framework 8 slot standard cabinets and 32 passage high-speed data acquisition cards for use, the LabVIEW language that Software Development Platform selects the said firm to provide, signals collecting adopts Hall current sensor to realize.
The process flow diagram of pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and method for diagnosing faults as shown in Figure 1, comprising: 1. data acquisition; 2. digital filtering adopts particular algorithm that data are carried out filtering and pre-service; 3. ask for extreme point, the local maximum minimal point in the search one dimension scalar array carries out the broken line processing to waveform; 4. extreme point merges and unimodal elimination processing, and unnecessary end points merges after the doubling linearize; 5. divide to promote, transmit, keep three kinds of situation identification waveform character points, check size of current and fluctuating range; 6. differentiate and fault diagnosis step preface, actuation time, lifting coil current waveform, transfer line loop current waveform, holding coil current waveform are carried out analyzing and processing, form the operating point array, judge whether to be normal step preface, finish single CRDM step preface inspection; 7. the consistance to 4 cluster actions of a son group compares, and carries out fault diagnosis according to the deviation size, finishes the single step analysis of son group waveform; 8. gather waveform demonstration, valid data storage, action record storage.
1. data acquisition
Be changed to 2000Hz in the frequency acquisition, buffer area 4000; Each reading of data amount can be provided with in master routine.Can in master routine, stop or restarting collection.Sky array and failure message are returned in output again after data filtering is amplified when collection fault (handling untimely as data) is arranged.
2. data filtering and pre-service
Remove the single-point burr that occurs in the data acquisition.Disposal route: according to certain a bit and before and after it relation of some handle, substitute currency with predicted value when being judged as exceptional value.As shown in Figure 2, establishing the c point is pending point, and a, b are continuous 2 points in its front, and d, e, f are continuous 3 points in its back, judges according to the magnitude relationship of some before and after c and its whether it is unusual.If min{|c-b|, | c-d|}＞2 * max{|a-b|, | b-d|, | d-e|} then claim " c departs from a, b, d, e ", three kinds of situations of calculating value distribution: relatively | a-b|, | b-d|, the little usefulness of the latter (b+d)/2, when rising as the former small form with b+ (b-a)/2, during decline with b+ (b-a) (fall increase gradually always) because ascensional range always reduces gradually.The c point is a singular point obviously among Fig. 2, and c ' point is its predicted value.
Data filtering and pretreated process flow diagram, as shown in Figure 3.
(1) putting preceding two point prediction values is preceding 6 averages;
(2) since the 3rd data, up to third from the bottom point, following processing is carried out in circulation:
Judge whether c ≈ 0 or be lasting rising or intermediate point for continuing to descend? be, then do not deal with; , then do not judge whether to be one of following situation:
Be then to get b, d mean value is the c predicted value;
1. fall greatly and fall: the peak that occurs in the decline process
2. fall for a short time and fall: the ebb that occurs in the decline process
3. rise and rise higher: occur the peak during rising
4. rise little liter: occur ebb during rising
5. fall and rise higher: the peak appears in the commentaries on classics rising portion that descends
6. rise and fall for a short time: rising to change ebb occurs when descending
Not, judge whether that then c departs from a, b, d, f? be, then predict the c value according to a, b, d, f; Deny, then judge whether that c departs from a, b, e, f?
Be, then according to a, b, e, f prediction c value; Not, judge whether that then c departs from a, b, d, e?
Be, then predict the c value according to a, b, d, e; Not, judge whether that then c is for going up the summit unusually? c is that last summit and c-b＞b-a think that then c goes up the summit unusually,
Be, then the c value changes b+ (b-a)/2 into; Deny, then judge whether that c is for descending the summit unusually? c is that following summit and b-c＞a-b think that then c descends the summit unusually,
Be, then the c value changes 2b-a into; , then the c value is not constant;
(3) 2 of postpones are the predicted value of 4 of its fronts.
Fig. 4 is data filtering and effects of pretreatment figure.
3. ask for extreme point, waveform is carried out the broken line processing
Local maximum minimal point in the search one dimension scalar array.
Currency is greater than last value or approximate last value and be in local ascent stage, confirms when minimal point is unconfirmed to upgrade unique point by minimal point, and short-term trend changes rising into, and minimal point changes into confirms that maximal point changes into unconfirmed;
Otherwise then currency upgrades unique point less than last value or approximate last value and be in and local confirm maximal point when the decline stage, maximal point was unconfirmed, and short-term trend changes decline into, and maximal point changes into to be confirmed, minimal point changes into unconfirmed.
The design sketch that waveform broken lineization, merging are handled, as shown in Figure 5.The top be original waveform, middle is waveform after the broken line processing.
4. extreme point merges and unimodal elimination processing
Eliminate the improper end points that Yin Maoci produces, remove unimodal that the broad noise forms.Rudimentary algorithm is as follows:
1. as minimal point less than next minimal point, delete next maximal point, next minimal point.
2. as minimal point 〉=next minimal point, deletion itself reaches next maximal point.
3. as maximal point less than next maximal point, deletion itself and next minimal point.
4. as maximal point 〉=next maximal point, delete next minimal point, maximal point.
The design sketch that broken lineization, merging are handled, as shown in Figure 5.It is following to be waveform after the merging of broken line end points.
5. divide lifting (LC), transmit (MC), keep (SC) three kinds of situations that waveform character is carried out identification, check size of current and fluctuating range.
Set the criterion that triggers each state exchange, these criterions are the keys that guarantee good waveform identification effect.
5.1 promote coil current waveform finite state machine state be: (1) original state; (2) zero current platform; (3) rising 0-40; (4) adhesive platform; (5) big electric current platform; (6) decline 40-16; (7) little electric current platform; (8) decline 16-0; (9) stop.
Ask for and promote starting point, promote the adhesive starting point, promote the adhesive operating point, promote the large platform terminal point, promote the chain-wales starting point, promote the chain-wales terminal point, promote 7 operating points of terminal point.
Promote coil current waveform character identification state machine, as shown in Figure 6.
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 1A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-40; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 1A, enter (9) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-40;
(3) rising 0-40 when certain minimal point≤0, enters (2) zero current platform; When the average of three maximal points of certain maximal point and back differ＜during 1A, enter (4) adhesive platform;
(4) adhesive platform, three the maximal point averages in back enter (5) big electric current platform than more than three the big 0.5A of maximal point average in front; When behind adhesive platform 100ms, not finding operating point yet, enter (5) big electric current platform; Certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform;
(5) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (6) decline 40-16;
(6) decline 40-16, half of certain minimal point-back minimal point≤back increasing degree enters (7) little electric current platform;
(7) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (8) decline 16-0;
(8) decline 16-0, certain minimal point or thereafter maximal point be about 0, enter (9) and stop;
(9) stop.
5.2 transfer line loop current waveform finite state machine state is: (1) original state; (2) zero current platform; (3) rising 0-8; (4) descend before the adhesive; (5) rise after the adhesive; (6) big electric current platform; (7) decline 8-4; (8) little electric current platform; (9) decline 4-0; (10) stop.
Ask for and transmit starting point, transmit the adhesive starting point, transmit the adhesive operating point, transmit the large platform starting point, transmit the large platform terminal point, transmit the chain-wales starting point, transmit the chain-wales terminal point, transmit 8 operating points of terminal point.
Transfer line loop current waveform character identification state machine, as shown in Figure 7.
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 0.5A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-8; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 0.5A, enter (10) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-8;
(3) rising 0-8 when certain minimal point≤0, enters (2) zero current platform; When certain maximal point＞last maximal point, simultaneously＞next maximal point-0.1A, enter the preceding decline of (4) adhesive;
(4) descend before the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; When behind adhesive platform 160ms, not finding operating point yet, rise after entering (5) adhesive; Several minimal points before and after a certain minimal point is significantly less than rise after entering (5) adhesive;
(5) rise after the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; A certain maximal point is greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (6) big electric current platform.
(6) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (7) decline 8-4;
(7) decline 8-4, certain minimal point＜1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (10) and stop; , certain minimal point＞1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (8) little electric current platform;
(8) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (9) decline 4-0;
(9) decline 4-0, certain minimal point or thereafter maximal point be about 0, enter (10) and stop;
(10) stop.
5.3 holding coil current waveform finite state machine state is: (1) original state; (2) little electric current keeps; (3) rising 4-8; (4) big electric current keeps; (5) decline 8-0; (6) zero current platform; (7) rising 0-8; (8) descend before the adhesive; (9) rise after the adhesive; (10) big electric current platform; (11) decline 8-4; (12) stop.
Ask for keep starting point, keep before the platform starting point, keep before the platform terminal point, keep discharging return-to-zero, keep heavily adhesive starting point, keep the action starting point, keep the adhesive operating point, keep the large platform starting point, keep the large platform terminal point, keep 10 operating points of terminal point.
Holding coil Current Waveform Characteristics identification state machine, as shown in Figure 8.
(1) original state, the relative wave amplitude of certain minimal point are between 0.2-0.5, and be very little with back minimal point relative deviation, enters (2) little electric current and keep; Two maximal points in certain minimal point back significantly rise and current point＞0.8 times increasing degree continuously, enter (3) rising 4-8; Certain minimal point is about 0, enters (6) zero current platform; Two maximal points in certain minimal point back significantly rise and current point≤0.8 times increasing degree continuously, enter (7) rising 0-8; Two the minimal point＞6A in front, current point＜5A, and back two minimal points＞3.5A enter (12) and stop;
(2) little electric current keeps, and two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 4-8; Last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(3) rising 4-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (4) big electric current and keep; A certain minimal point is about 0, enters (6) zero current platform;
(4) big electric current keeps, and the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enters (5) decline 8-0;
(5) decline 8-0, last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(6) zero current platform, two maximal points in certain minimal point back significantly rise continuously, enter (7) rising 0-8;
(7) rising 0-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, descends before entering (8) adhesive;
(8) descend before the adhesive, behind adhesive platform 100ms, do not find operating point yet, enter (10) big electric current platform; Several minimal points before and after a certain minimal point is significantly less than rise after entering (9) adhesive;
(9) rise after the adhesive, differ half, enter (10) big electric current platform less than the range of decrease with three maximal points in back;
(10) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (11) decline 8-4;
(11) decline 8-4,0.5 times of minimal point-back minimal point≤back increasing degree enters (12) and stops;
(12) stop.
6. differentiate and fault diagnosis step preface, actuation time
Promote, transmit, keep three coil current waveforms to carry out analyzing and processing to a cluster, form the operating point array, judge whether to be normal step preface, in normal range, whether each operating point context meets the demands the difference of judging operating point and standard operation point.
The operating point array comprises 30 data, wherein 7 promote operating point, 6 transmit operating point, 10 keep operating point, 5 actuation time data, 2 valid data terminal information.
7. the consistance to 4 cluster actions of a son group compares
The consistance of 4 cluster actions compares in the antithetical phrase group.The operating point consistance establishes relatively that deviation is big, the serious two kinds of threshold values of deviation.
8. gather waveform demonstration, valid data storage, action record storage
Waveform after demonstration original waveform and the processing; show step round-robin information actuation time with form; as judge that waveform is unusual or comprise effective action and then preserve waveform, action message, actuation time are deposited in the database in display abnormality or failure message in monitoring image.
Utilize message passing mechanism to realize that nine son group parallel acquisitions handle, fault, waveform unusual or that comprise effective action are preserved, shown.
Action message, actuation time and the failure logging information of gathering gained waveform and generation send on the bus by message passing mechanism, show in real time and with actuation time information, failure logging information deposit database in.
This method can realize online high speed acquisition, processing, the storage of 144 analog signalses, can satisfy the requirement of CRDM on-line monitoring and fault diagnosis.9 son group operations simultaneously, per 1 data processing, waveform storage time in second is less than 500 milliseconds; Historical record is checked, the off-line data analysis does not influence on-line monitoring; CRDM action, current anomaly fluctuation etc. are to the equal timely record of the significant waveform of ex-post analysis.
Claims (2)
1. the inline diagnosis of a pressurized water reactor nuclear power station control bar driving mechanism and method for diagnosing faults may further comprise the steps:
(1) data acquisition;
(2) digital filtering;
(1) establishing the c point is pending point, and a, b are continuous 2 points in its front, and d, e, f are continuous 3 points in its back, judges according to the magnitude relationship of some before and after c and its whether it is unusual;
If min{|c-b|, | c-d|}＞2 * max{|a-b|, | b-d|, | d-e|} then claims c to depart from a, b, d, e, three kinds of situations of calculating value distribution: relatively | a-b|, | b-d|, the little usefulness of the latter (b+d)/2 with b+ (b-a)/2, is used b+ (b-a) when rising as the former small form during decline;
(2) putting preceding two point prediction values is preceding 6 averages;
(3) since the 3rd data, up to third from the bottom point, following processing is carried out in circulation:
Judge whether c ≈ 0 or be lasting rising or intermediate point for continuing to descend? be, then do not deal with; , then do not judge whether to be one of following situation:
Be then to get b, d mean value is the c predicted value;
1. fall greatly and fall: the peak that occurs in the decline process;
2. fall for a short time and fall: the ebb that occurs in the decline process;
3. rise and rise higher: occur the peak during rising;
4. rise little liter: occur ebb during rising;
5. fall and rise higher: the peak appears in the commentaries on classics rising portion that descends;
6. rise and fall for a short time: rising to change ebb occurs when descending;
Not, judge whether that then c departs from a, b, d, f? be, then according to a, b, d, f prediction c value; Not, judge whether that then c departs from a, b, e, f?
Be, then according to a, b, e, f prediction c value; Not, judge whether that then c departs from a, b, d, e?
Be, then according to a, b, d, e prediction c value; Not, judge whether that then c is for going up the summit unusually? c is that last summit and c-b＞b-a think that then c goes up the summit unusually;
Be that then the c value changes b+ (b-a)/2 into; Not, judge whether that then c is for descending the summit unusually? c is that following summit and b-c＞a-b think that then c descends the summit unusually;
Be that then the c value changes 2b-a into; Not, then the c value is constant;
(4) 2 of postpones are the predicted value of 4 of its fronts;
(3) ask for extreme point, waveform is carried out the broken line processing
Local maximum minimal point in the search one dimension scalar array, currency is greater than last value or approximate last value and be in local ascent stage, confirms minimal point when minimal point is unconfirmed, upgrade unique point, short-term trend changes rising into, and minimal point changes into confirms that maximal point changes into unconfirmed; Otherwise then currency upgrades unique point less than last value or approximate last value and be in and local confirm maximal point when the decline stage, maximal point was unconfirmed, and short-term trend changes decline into, and maximal point changes into to be confirmed, it is unconfirmed that minimal point changes into;
(4) extreme point merges and unimodal elimination processing
(1) if certain minimal point, is then deleted next maximal point, next minimal point less than next minimal point;
(2) if certain minimal point 〉=next minimal point, then deletion itself and next maximal point;
(3) if certain maximal point less than next maximal point, then deletion itself and next minimal point;
(4) if certain maximal point 〉=next maximal point is then deleted next minimal point, maximal point;
(5) divide lifting, transmission, three kinds of situations identification of maintenance waveform character points, check size of current and fluctuating range
Adopt the finite state machine pattern to promoting the method that coil current waveform character point is discerned, may further comprise the steps:
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 1A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-40; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 1A, enter (9) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-40;
(3) rising 0-40 when certain minimal point≤0, enters (2) zero current platform; When the average of three maximal points of certain maximal point and back differ＜during 1A, enter (4) adhesive platform;
(4) adhesive platform, three the maximal point averages in back enter (5) big electric current platform than more than three the big 0.5A of maximal point average in front; When behind adhesive platform 100ms, not finding operating point yet, enter (5) big electric current platform; Certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform;
(5) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (6) decline 40-16;
(6) decline 40-16, half of certain minimal point-back minimal point≤back increasing degree enters (7) little electric current platform;
(7) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (8) decline 16-0;
(8) decline 16-0, certain minimal point or thereafter maximal point be about 0, enter (9) and stop;
(9) stop;
Adopt the finite state machine pattern to the method that transfer line loop current ripple carries out unique point identification, may further comprise the steps:
(1) original state, current minimal point or back maximal point are about 0 and two minimal points in front during all less than 0.5A, enter (2) zero current platform; When two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 0-8; Current minimal point or back maximal point be about 0 and two minimal points in front have one greater than 0.5A, enter (10) and stop;
(2) zero current platform when two maximal points in certain minimal point back significantly rise continuously, enters (3) rising 0-8;
(3) rising 0-8 when certain minimal point≤0, enters (2) zero current platform; When certain maximal point＞last maximal point, simultaneously＞next maximal point-0.1A, enter the preceding decline of (4) adhesive;
(4) descend before the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; When behind adhesive platform 160ms, not finding operating point yet, rise after entering (5) adhesive; Several minimal points before and after a certain minimal point is significantly less than rise after entering (5) adhesive;
(5) rise after the adhesive, certain minimal point or thereafter maximal point be about at 0 o'clock, enter (2) zero current platform; A certain maximal point is greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (6) big electric current platform;
(6) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (7) decline 8-4;
(7) decline 8-4, certain minimal point＜1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (10) and stop; , certain minimal point＞1A and its deduct 0.2 times of back minimal point≤back increasing degree, enter (8) little electric current platform;
(8) little electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (9) decline 4-0;
(9) decline 4-0, certain minimal point or thereafter maximal point be about 0, enter (10) and stop;
(10) stop;
Adopt the finite state machine pattern to the method that the holding coil current waveform carries out unique point identification, may further comprise the steps:
(1) original state, the relative wave amplitude of certain minimal point are between 0.2-0.5, and be very little with back minimal point relative deviation, enters (2) little electric current and keep; Two maximal points in certain minimal point back significantly rise and current point＞0.8 times increasing degree continuously, enter (3) rising 4-8; Certain minimal point is about 0, enters (6) zero current platform; Two maximal points in certain minimal point back significantly rise and current point≤0.8 times increasing degree continuously, enter (7) rising 0-8; Two the minimal point＞6A in front, current point＜5A, and back two minimal points＞3.5A enter (12) and stop;
(2) little electric current keeps, and two maximal points in certain minimal point back significantly rise continuously, enter (3) rising 4-8; Last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(3) rising 4-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, enters (4) big electric current and keep; A certain minimal point is about 0, enters (6) zero current platform;
(4) big electric current keeps, and the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enters (5) decline 8-0;
(5) decline 8-0, last minimal point or back one maximal point are about 0, enter (6) zero current platform;
(6) zero current platform, two maximal points in certain minimal point back significantly rise continuously, enter (7) rising 0-8;
(7) rising 0-8, certain maximal point be greater than last maximal point, differs less than next range of decrease 0.3 times with two maximal points in back, descends before entering (8) adhesive;
(8) descend before the adhesive, behind adhesive platform 100ms, do not find operating point yet, enter
(10) big electric current platform; Several minimal points before and after a certain minimal point is significantly less than rise after entering (9) adhesive;
(9) rise after the adhesive, differ half, enter (10) big electric current platform less than the range of decrease with three maximal points in back;
(10) big electric current platform, the minimum minimal point of the big electric current platform in certain minimal point place obviously greater than second and third maximal point thereafter, enter (11) decline 8-4;
(11) decline 8-4,0.5 times of minimal point one back minimal point≤back increasing degree enters (12) and stops;
(12) stop;
(6) differentiate and fault diagnosis step preface, actuation time, lifting coil current waveform, transfer line loop current waveform, holding coil current waveform are carried out analyzing and processing, form the operating point array, judge whether to be normal step preface, finish single CRDM step preface inspection;
(7) consistance to 4 cluster actions of a son group compares, and carries out fault diagnosis according to the deviation size, finishes the single step analysis of son group waveform;
(8) gather waveform demonstration, valid data storage, action record storage.
2. the inline diagnosis of pressurized water reactor nuclear power station control bar driving mechanism as claimed in claim 1 and method for diagnosing faults; it is characterized in that: adopt virtual instrument technique; utilize message passing mechanism to realize the processing of a plurality of son group parallel acquisitions; action message, actuation time recorded information, the failure logging information of gathering gained waveform and generation send on the bus by message passing mechanism; show in real time; simultaneously with actuation time information, failure logging information deposit database in, adopt the form of file to preserve fault, waveform unusual or that comprise effective action.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610152151A CN101067976B (en) | 2006-09-15 | 2006-09-15 | Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610152151A CN101067976B (en) | 2006-09-15 | 2006-09-15 | Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101067976A CN101067976A (en) | 2007-11-07 |
| CN101067976B true CN101067976B (en) | 2010-05-19 |
Family
ID=38880487
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200610152151A Active CN101067976B (en) | 2006-09-15 | 2006-09-15 | Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101067976B (en) |
Families Citing this family (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101847452B (en) * | 2009-08-31 | 2012-04-18 | 中广核工程有限公司 | Method and system for diagnosing first fault of pressurized-water reactor nuclear power plant |
| CN101840737B (en) * | 2009-09-02 | 2012-10-31 | 中广核工程有限公司 | Digital rod position control system and method |
| CN102062838B (en) * | 2010-11-23 | 2012-10-10 | 肖闽进 | Real-time data acquisition device of 220kV main transformer oil pump fan motor cluster |
| CN102708935B (en) * | 2012-03-23 | 2015-10-14 | 中广核工程有限公司 | Nuclear Power Plant Simulation C&P systems |
| CN202735877U (en) * | 2012-05-17 | 2013-02-13 | 上海核工程研究设计院 | Nuclear power plant digital control rod control system based on field programmable gate array (FPGA) and digital signal processing (DSP) controller |
| CN104332196A (en) * | 2013-07-22 | 2015-02-04 | 中国核动力研究设计院 | Method for monitoring reactor control-rod driving-mechanism motion point |
| CN104330611A (en) * | 2013-07-22 | 2015-02-04 | 中国核动力研究设计院 | Method for monitoring reactor control rod drive mechanism coil current amplitude |
| CN104392752B (en) * | 2014-10-13 | 2016-11-30 | 中国科学院合肥物质科学研究院 | The nuclear reactor fault diagnosis of a kind of real-time online and monitoring system |
| CN105070334B (en) * | 2015-06-24 | 2019-01-25 | 中国核电工程有限公司 | A method of judging that control rod drive mechanism acts based on curent change |
| CN105551543B (en) * | 2016-01-15 | 2018-04-03 | 中广核工程有限公司 | A kind of rod control and rod position system and its method for diagnosing faults for nuclear power station |
| CN108548952A (en) * | 2018-05-11 | 2018-09-18 | 中国核动力研究设计院 | A kind of reactor control rod driving mechanism current monitoring system |
| CN108646084A (en) * | 2018-05-11 | 2018-10-12 | 中国核动力研究设计院 | The method for monitoring the oscillation of reactor control rod driving mechanism coil current rising edge |
| CN108957317B (en) * | 2018-06-06 | 2021-01-26 | 广东核电合营有限公司 | Method and device for detecting bearing bush of generator set of pressurized water reactor nuclear power station, computer and storage medium |
| CN109033996B (en) * | 2018-07-03 | 2022-02-11 | 运易通科技有限公司 | Waveform analysis method and system based on motion data |
| CN110415844B (en) * | 2019-06-21 | 2021-06-01 | 杭州安脉盛智能技术有限公司 | Performance evaluation method and system for control rod driving mechanism based on action key time sequence |
| CN110570958B (en) * | 2019-08-28 | 2021-08-17 | 中广核研究院有限公司 | Method for monitoring hook action of control rod driving mechanism of nuclear power station and electronic equipment |
| CN117373709B (en) * | 2023-09-21 | 2024-04-02 | 中山大学 | Reactor core measurement method based on model order reduction and mixed data assimilation |
-
2006
- 2006-09-15 CN CN200610152151A patent/CN101067976B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 昌正科.核电站棒控棒位设备故障诊断技术研究与实践.南京工业大学硕士论文.2005,32-33. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101067976A (en) | 2007-11-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101067976B (en) | Pressurized water reactor nuclear power station control bar driving mechanism on-line monitoring and fault diagnosing method | |
| CN103487749A (en) | On-line monitoring and diagnosing system and method for mechanical state of high-voltage circuit breaker | |
| CN102147268B (en) | Integral panoramic data recording and analyzing system | |
| CN102102629A (en) | On-line data acquisition and analysis device of wind generating set | |
| CN101149416A (en) | Power cable insulation state monitoring and life span management system | |
| CN112692646A (en) | Intelligent assessment method and device for tool wear state | |
| CN107891199A (en) | Spark discharge condition checkout gear and recognition methods | |
| CN105718958A (en) | Breaker fault diagnosis method based on linear discriminant analysis and support vector machine | |
| CN102842344B (en) | Method for testing EEPROM (electrically erasable programmable read-only memory) read-write cycle times | |
| CN113762182A (en) | Milling cutter wear state prediction method and device based on depth network self-adaptation | |
| CN110968703B (en) | Method and system for constructing abnormal metering point knowledge base based on LSTM end-to-end extraction algorithm | |
| CN101254528A (en) | Bale discharging slag testing automatic control sytem | |
| CN112305388A (en) | On-line monitoring and diagnosing method for partial discharge fault of generator stator winding insulation | |
| CN113485244A (en) | Numerical control machine tool control system and method based on cutter wear prediction | |
| CN202817128U (en) | Storage battery detecting and matching device | |
| KR102556093B1 (en) | Reward generation method to reduce peak load of electric power and action control apparatus performing the same method | |
| CN112415350A (en) | Online detection method and device for series arc fault | |
| CN201754469U (en) | Recording and analyzing device for operation and failure of switch cabinet | |
| CN107328994A (en) | Insulation Resistance of Transformer experimental rig and method | |
| Wu et al. | An improved LSTM neural network with uncertainty to predict remaining useful life | |
| CN114355184A (en) | High-voltage circuit breaker state monitoring and early warning system and method based on online learning | |
| CN112561192A (en) | AI artificial intelligence based power load prediction system | |
| CN106528401B (en) | The tripping test method and device of control system based on redundancy structure | |
| CN202110191U (en) | Portable locomotive speedometer detection equipment | |
| CN110968447A (en) | Server host inspection system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |