CN104571084A - Further diagnosis method and device for root causes of failure of main fan set - Google Patents
Further diagnosis method and device for root causes of failure of main fan set Download PDFInfo
- Publication number
- CN104571084A CN104571084A CN201410771549.0A CN201410771549A CN104571084A CN 104571084 A CN104571084 A CN 104571084A CN 201410771549 A CN201410771549 A CN 201410771549A CN 104571084 A CN104571084 A CN 104571084A
- Authority
- CN
- China
- Prior art keywords
- fault
- fault database
- transition
- transition rule
- database
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0275—Fault isolation and identification, e.g. classify fault; estimate cause or root of failure
- G05B23/0278—Qualitative, e.g. if-then rules; Fuzzy logic; Lookup tables; Symptomatic search; FMEA
Abstract
The invention provides a further diagnosis method and a further diagnosis device for root causes of a failure of a main fan set. The method comprises the steps of constructing a fuzzy failure Petri net diagram according to a mutual relation among all failure libraries of a main fan set system; determining an initial credibility set of all the failure libraries, a failure event weight set of all the failure libraries, a failure rate set of all protection layer libraries, a threshold value set of all transition rules and a credibility set of the transition rules according to expert experience data and on-site actual statistical data; searching a transmission path of the failure library with the maximum credibility on each layer from the fuzzy failure Petri net diagram according to the information; taking the failure libraries at the start points of the determined transmission paths as the root causes of the failure in main fan set equipment. According to the further diagnosis method, the technical problem that the root causes of the failure in the main fan set equipment cannot be accurately and effectively detected in the prior art is solved, and the technical effect of effectively improving the detection accuracy and the detection efficiency is achieved.
Description
Technical field
The present invention relates to fault diagnosis technology field, particularly a kind of Main air compressor fault rootstock degree of depth diagnostic method and device.
Background technology
As the key equipment of fluidized catalytic cracker, the Energy recovery machine set (abbreviation Main air compressor) of main air blower-flue gas turbine expander can burn required oxygen for regenerative system provides, thus reaches the object recovered energy.Main air compressor plays significant role in energy-saving field, economic benefits, but very responsive to fault.Under Main air compressor is operated in high temperature, environment at a high speed, and be subject to the erosion of high speed catalyst dust air-flow, cause there is a large amount of sudden, coupling and dependent failure.In order to avoid the generation of major accident, guarantee the safety in production of fluidized catalytic cracker, research Main air compressor being carried out to fault diagnosis technology is very necessary.
To Main air compressor carry out fault diagnosis not only can the Timeliness coverage fault symptom of a trend, suppress the development of fault, the cyclic utilization rate of resource can also be improved, obtain larger economic benefit, promote the development of low-carbon economy.
At present, common Main air compressor method for diagnosing faults mainly contains: Bayesian network, based on signed digraph method, artificial neural network method etc., but these methods have respective shortcoming in practical engineering application.Such as: the reasoning algorithm of Bayesian network is too complicated, the fault sample that the training need of artificial neural network is more, the order of accuarcy based on signed digraph depends on the understanding of modeling personnel to system.
Due to the existence of the problems referred to above, current Main air compressor method for diagnosing faults all cannot be applied well in Practical Project field, for the problems referred to above, not yet proposes effective solution at present.
Summary of the invention
Embodiments provide a kind of Main air compressor fault rootstock degree of depth diagnostic method, to solve the technical matters that cannot detect the root primordium of Main air compressor device fails in prior art accurately and effectively.The method comprises:
According to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
According to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
According to described each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Using the starting point fault database in the direction of fault propagation determined as the root primordium of Main air compressor device fails.
In one embodiment, according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure, comprising:
Determine to set up the tuple needed for described fuzzy fault Petri network figure, wherein, described tuple comprises: P is gathered in P that fault database is gathered, protective seam storehouse
iPL, transition rule set T, fault database event of failure weight vector set w, the threshold value set λ of transition rule, transition rule confidence level set μ, change after TL that fault database is gathered, each fault database initially change confidence level set M
0, each protective seam storehouse crash rate set L and the regular collection R of fuzzy production;
Propagate with the rule of the fuzzy production in the regular collection of described fuzzy production, in conjunction with other tuple information in described tuple, set up represent Main air compressor system each fault database between the described fuzzy fault Petri network figure of mutual relationship.
In one embodiment; according to described each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule; find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation, comprising:
Obtain initial failure P that storehouse is gathered;
Set up with described initial failure storehouse gather transition rule set T corresponding to P;
Set up triggered change after TL that fault database is gathered;
Determine fault database in described Main air compressor system gather relation in P and transition rule set T between each transition rule, determine current enable transition rule;
Trigger all current enable transition rules, try to achieve transition firing sequence;
Confidence level set M is once changed according to before each fault database institute
a-1, ask for current each fault database institute confidence level set M
a, wherein, occur causing rear collection fault database each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, wherein, a represents transition number of times;
The fault database set that obtains after transition is substituted in described fuzzy fault Petri network figure as parameter, find out the maximum fault database of each later reliability as the direction of fault propagation.
In one embodiment, determine current enable transition rule, comprising:
By transition before collect fault database institute to the actual confidence value of transition rule be not less than transition triggering threshold value, and this transition rule do not belong to triggered transition rule set then by this transition rule, be defined as current enable transition rule.
In one embodiment, current each fault database institute confidence level set M is asked for
a, comprising:
Determine whether matcoveredn intervention effect;
When matcoveredn intervention effect, using each protective seam storehouse crash rate set in each protective seam storehouse crash rate as according to one of ask for the confidence level set of current each fault database institute.
The embodiment of the present invention additionally provides a kind of Main air compressor fault rootstock degree of depth diagnostic device, to solve the technical matters that cannot detect the root primordium of Main air compressor device fails in prior art accurately and effectively.This device comprises:
Fuzzy fault Petri network figure sets up unit, for according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
Data capture unit, for according to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
Travel path searches unit, for according to described each fault database initial trusted degree set, each fault database the set of event of failure weights and each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Root primordium determining unit, for the starting point fault database of travel path that will determine as the root primordium of Main air compressor device fails.
In one embodiment, described fuzzy fault Petri network figure sets up unit and comprises:
Tuple determination module, set up the tuple needed for described fuzzy fault Petri network figure for determining, wherein, described tuple comprises: P is gathered in P that fault database is gathered, protective seam storehouse
iPL, transition rule set T, fault database event of failure weight vector set w, the threshold value set λ of transition rule, transition rule confidence level set μ, change after TL that fault database is gathered, each fault database initially change confidence level set M
0, each protective seam storehouse crash rate set L and the regular collection R of fuzzy production;
Net figure sets up module, for propagating with the rule of the fuzzy production in the regular collection of described fuzzy production, in conjunction with other tuple information in described tuple, set up represent Main air compressor system each fault database between the described fuzzy fault Petri network figure of mutual relationship.
In one embodiment, described travel path is searched unit and is comprised:
Acquisition module, for obtaining initial failure P that storehouse is gathered;
First sets up module, for set up with described initial failure storehouse gather transition rule set T corresponding to P;
Second sets up module, for set up triggered change after TL that fault database is gathered;
Enable transition rule determination module, for determine fault database in described Main air compressor system gather relation in P and transition rule set T between each transition rule, determine current enable transition rule;
Transition trigger module, for triggering all current enable transition rules, tries to achieve transition firing sequence;
Ask for module, for once changing confidence level set M according to before each fault database institute
a-1, ask for current each fault database institute confidence level set M
a, wherein, occur causing rear collection fault database each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, wherein, a represents transition number of times;
Travel path searches module, for the fault database set that obtains after transition is substituted in described fuzzy fault Petri network figure as parameter, find out the maximum fault database of each later reliability as the direction of fault propagation.
In one embodiment, described enable transition rule determination module will be specifically for collecting fault database institute and to be not less than the actual confidence value of transition rule the threshold value of transition triggering before transition, and this transition rule does not belong to and triggered transition rule set and then this transition rule is defined as current enable transition rule.
In one embodiment; described module of asking for is specifically for determining whether matcoveredn intervention effect; and when defining protective seam intervention effect, using each protective seam storehouse crash rate set in each protective seam storehouse crash rate as according to one of ask for the confidence level set of current each fault database institute.
In embodiments of the present invention, by each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure, and according to expertise data and on-the-spot actual count data, define each fault database the set of initial trusted degree, each fault database the set of event of failure weights, with each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, thus it can be used as parameter to substitute into above-mentioned fuzzy fault Petri network figure, can effectively find out the maximum fault database of each later reliability as the direction of fault propagation, to determine the root primordium for Main air compressor device fails, thus solve the technical matters that cannot detect the root primordium of Main air compressor device fails in prior art accurately and effectively, reach the technique effect effectively improving accuracy in detection and detection efficiency.Further, because introduce protective seam storehouse institute, make testing result more accurate.
Accompanying drawing explanation
Accompanying drawing described herein is used to provide a further understanding of the present invention, forms a application's part, does not form limitation of the invention.In the accompanying drawings:
Fig. 1 is the method flow diagram of the Main air compressor fault rootstock degree of depth diagnostic method according to the embodiment of the present invention;
Fig. 2 is the fuzzy fault Petri network figure schematic diagram according to the embodiment of the present invention;
Fig. 3 is the structured flowchart of the Main air compressor fault rootstock degree of depth diagnostic device according to the embodiment of the present invention.
Embodiment
For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with embodiment and accompanying drawing, the present invention is described in further details.At this, exemplary embodiment of the present invention and illustrating for explaining the present invention, but not as a limitation of the invention.
For the deficiency of common Main air compressor method for diagnosing faults; in embodiments of the present invention; propose a kind of Main air compressor fault rootstock degree of depth diagnostic method; the Main air compressor multifactor relevant failure root degree of depth diagnosis of protective seam intervention is considered in this detection method; concrete; as shown in Figure 1, comprise the following steps:
Step 101: according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
Step 102: according to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree set, each fault database the set of event of failure weights and each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
Step 103: according to described each fault database the set of initial transition confidence level, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Step 104: using the starting point fault database of travel path determined as the root primordium of Main air compressor device fails.
Namely; by analyzing the fault mode of each major equipment of Main air compressor and cause-effect relationship; determine each fault database protective seam storehouse institute in one's power; FFPN (the Fuzzy Fault Petri Net) model of the multifactor relevant failure fault propagation of a kind of Main air compressor and protective seam intervention is set up based on fuzzy fault Petri network, concrete:
FFPN net can be defined as 10 tuples: FFPN=(P, a P
iPL, T, w, λ, μ, TL, M
0, L, R), below each tuple in 10 tuples is specifically described:
1) P={p
1, p
2..., p
n, represent that limited fault storehouse gathered, p
k(k=1,2 ..., n) represent each fault database institute, n is fault database institute number;
2) P
iPL={ p
iPL1, p
iPL2..., p
iPLo, represent that limited protective seam storehouse gathered, p
iPLk(k=1,2 ... o) represent each protective seam storehouse institute, o is protective seam storehouse institute number;
3) T={t
1, t
2..., t
m, represent limited transition rule set, t
j(j=1,2 ... m) represent each transition rule, m represents transition rule number;
4) w=(w
1, w
2..., w
n)
t, represent fault database the set of event of failure weight vector, w
i(i=1,2 ... n) each weight vector is represented, i.e. input fault storehouse institute p
kto transition rule t
jinfluence degree;
5) λ={ λ
1, λ
2..., λ
m, represent the threshold value set of transition rule, λ
jrepresent transition rule t
jthe threshold value triggered, λ
j∈ [0,1], j=1,2 ..., m;
6) μ=diag (μ
1, μ
2..., μ
m), represent the confidence level set of transition rule, μ
jrepresent transition rule t
jconfidence level, μ
j∈ [0,1], j=1,2 ..., m;
7) M
a={ m
a(p
1), m
a(p
2) ..., m
a(p
n), represent the fault database institute confidence level set after generation a time transition, a is transition number of times, as a=0, and M
0={ m
0(p
1), m
0(p
2) ..., m
0(p
n) represent that fault database initially changes confidence level set, m
0(p
i) be fault database institute p
iinitial trusted degree, m
0(p
i) ∈ (0,1], i=1,2 ..., n, fault database the set of initial transition confidence level can according to observing history data and measuring acquisition;
8) TL, represents that the fault database after transition generation gathered, and time initial, this set is empty,
if collection fault database institute p before it
imeet m
a(p
i) * w
i>=λ
jand
then claim transition rule t
jbe enable (i=1,2 ..., n, j=1,2 ..., m), namely the front collection fault database changed to the actual confidence value of transition rule be not less than transition rule trigger threshold value, and this transition rule do not belong to triggered transition set, then this transition rule is enable, transition rule t
j(j=1,2 ..., after m) occurring, the fault database institute set expression after transition occur is: TL=TL+{t
j;
9) L={L
1, L
2..., L
o, represent each protective seam storehouse crash rate set, L
k(k=1,2 ... o) represent each protective seam storehouse crash rate;
10) R={R
1, R
2..., represent the regular collection of fuzzy production, R
ifor the rule of i-th in R, then R
ibe defined as p
i→ p
k, wherein, p
iand p
kbefore and after being respectively, the front collection fault database of contact collects fault database institute in one's power afterwards, front collection fault database institute p
irepresent certain failure cause, rear collection fault database institute p
krepresentative is by p
ithe phenomenon of the failure that may cause, m
ifor p
ito the confidence level of this rule, μ
jfor the confidence level of this rule, ω
irepresent each weight vector, collection fault database institute p namely
ito transition rule t
jinfluence degree, and μ
j, m
i, ω
i∈ [0,1], p
i→ p
krepresent the travel path of fault, if unprotect layer intervention effect, when transition rule occurs, m
i× μ
j× ω
isize represent the actual confidence level of every rule, if m
i× μ
j× ω
imore close to 1, then R
imore credible; If matcoveredn intervention effect in fault propagation process, when transition rule occurs, m
i× μ
j× ω
i× L
isize represent the actual confidence level of every rule, if m
i× μ
j× ω
i× L
imore close to 1, then R
imore credible.
After establishing above-mentioned fuzzy fault Petri network, just can perform above-mentioned steps 102, namely basis is to the statistics of Main air compressor actual operating state and expertise, determines each fault database institute initial trusted degree set M of fuzzy fault Petri network
0, each fault database event of failure weights set w, the threshold value set λ of transition rule, the confidence level set μ of transition rule and each protective seam storehouse crash rate set L; then just can according to following steps determine the fault database that in fuzzy fault Petri network figure, each later reliability is maximum travel path, concrete comprises the following steps:
Step 1: set up with initial failure storehouse gather transition rule set T corresponding to P, foundation triggered change after TL that fault database is gathered, make TL=Φ;
Step 2: make a=1, wherein, a represents transition number of times, for the number of times of Mk system reasoning, the number of times namely changed;
Step 3: analyze fault database in Main air compressor system gather relation in P and transition rule set T between each transition rule, find current enable transition, get t
j∈ T, if collection fault database institute p before it
imeet m
a(p
i) * w
i>=λ
jand
(wherein, m
a(p
i) represent the rear fault database institute p of generation a time transition
iconfidence level, ω
irepresent each weight vector, λ
jrepresent transition rule t
jthe threshold value triggered), then claim transition rule t
jbe enable (i=1,2 ..., n, j=1,2 ..., m), namely transition rule front collection fault database the threshold value that transition trigger is not less than to the actual confidence value of transition rule, and this transition rule does not belong to and has triggered transition rule set, then this transition rule is enable;
Step 4: trigger all current enable transition rules, obtains transition firing sequence according to the enable rule of transition, if unprotect layer intervention effect, when transition occur, and m
a(p
i) × μ
j× ω
isize represent the actual confidence level of every bar transition rule, if m
a(p
i) × μ
j× ω
imore close to 1, then R
imore credible; If matcoveredn intervention effect in fault propagation process, then when transition occur, m
a(p
i) × μ
j× ω
i× L
isize represent the actual confidence level of every bar transition rule, if m
a(p
i) × μ
j× ω
i× L
imore close to 1, then R
imore credible; If there is t
j(one or more transition) can trigger (namely fault may occur), make TL=TL+{t
j, T=T-TL, wherein, m
a(p
i) be fault database institute p after generation a time transition
iconfidence level, μ
jrepresent transition rule t
jconfidence level, ω
irepresent each weight vector, L
krepresent each protective seam storehouse crash rate;
Step 5: by M
a-1ask for current each fault database institute confidence level set M
a, fault is propagated by Fuzzy Production Rule, cause rear collection fault database to occur each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, that is, m
a+1(p
i)=max{m
a(p
i) × μ
j× ω
i× L
i, and value assignment maximum for confidence level in each competitive events causing the fault database institute confidence level of generation to transition, all fault databases not occurring to change gathered to make P represent, P
irepresent occur transition fault database set, make P=P-P
i;
Step 6: repeated execution of steps 3 to 5, until P=Φ.
Finally, can each set obtained be substituting in fuzzy fault Petri network as parameter, find out the maximum fault database of each later reliability travel path, using this travel path as fault most probable travel path in systems in which, namely in system, need the weak link that emphasis is monitored.
In the above-described embodiments, have employed a kind of detection carrying out Main air compressor device fails reason considering the Main air compressor of protective seam intervention multifactor relevant failure root degree of depth diagnostic analysis method, on the basis that FTA and FMEA analyzes, determine fault mode and the cause-effect relationship of each major equipment of Main air compressor, and determine each fault database protective seam storehouse institute in one's power, then fuzzy fault Petri network modeling method is used to carry out fault diagnosis and reasoning, the production status that this method is more realistic, can be quicker by reasoning, determine the weak link of out of order possibility travel path and Main air compressor system more accurately, can provide accurate for the formulation of on-site maintenance plan and monitoring plan, eliminate fault rootstock timely and reliable theoretical foundation is provided, thus ensure that Main air compressor runs safely and reliably.
When specific implementation, when in system, fault database institute is more, the carrying out of conveniently reasoning, can carry out the simplification of Petri network illustraton of model according to some rules, such as:
1) find when analyzing Main air compressor equipment failure, Network Abnormal can be caused huge and complicated if embody whole fault in FFPN model, therefore in order to simplified model scale, some fault can be classified as a class, such as: gear fatigue wear (comprising: spot corrosion, the flank of tooth come off, surface crushes) and tooth face agglutination fault can all be classified as tooth surface abrasion fault;
2) probability of occurrence of some fault is very low and just can eliminate in Maintenance and Repair at ordinary times, and these faults can not be embodied in network model, such as: securing member looseness fault.
Below in conjunction with a specific embodiment, above-mentioned Main air compressor fault rootstock degree of depth diagnostic method is described, but it should be noted that this specific embodiment is only to better the present invention is described, do not form inappropriate limitation of the present invention.
In this example, with Main air compressor major equipment, take flue gas turbine expander as research object, according to the statistics of Main air compressor actual operating state and expertise, determine the relevant parameters of fuzzy fault Petri network, carry out the fault diagnosis of flue gas turbine expander equipment.
Wherein, each fault database implication and initially change confidence level, storehouse institute fault database the confidence value of event of failure weights, the implication of transition, threshold value and transition rule as shown in table 1 below; each protective seam storehouse implication and crash rate can be as shown in table 2 below; the illustraton of model of the fuzzy fault Petri network set up can be as shown in Figure 2; in table 1 below; it is not original state that "-" represents this state, and its value needs produced by collecting fault database before being correlated with.
Table 1
Table 2
| Parameter | Implication | Crash rate L |
| PIPL1 | Shaft vibration faults of monitoring system and personnel intervene inefficacy | 0.10 |
| PIPL2 | Axle temperature degree faults of monitoring system and personnel intervene inefficacy | 0.05 |
| PIPL3 | Axial translation faults of monitoring system and personnel intervene inefficacy | 0.05 |
| PIPL4 | Temperature alarm lost efficacy | 0.15 |
| PIPL5 | Pressure warning unit lost efficacy | 0.10 |
| PIPL6 | Emergency stopping system lost efficacy | 0.09 |
| PIPL7 | Temperature alarm lost efficacy | 0.15 |
| PIPL8 | Safety interlock system high temperature failure | 0.15 |
| PIPL9 | Pressure warning unit lost efficacy | 0.10 |
| PIPL10 | Personnel intervene inefficacy | 0.32 |
| PIPL11 | Pressure warning unit lost efficacy | 0.10 |
| PIPL12 | Temperature alarm lost efficacy | 0.15 |
| PIPL13 | Pressure warning unit lost efficacy | 0.20 |
| PIPL14 | Pressure transducer lost efficacy | 0.10 |
| PIPL15 | Emergency stopping system lost efficacy | 0.09 |
| PIPL16 | Pressure transducer lost efficacy | 0.10 |
| PIPL17 | Pressure transducer lost efficacy | 0.10 |
Reasoning is carried out to flue gas turbine expander fault, from all events, suppose that each event all exists the possibility broken down, namely given in table 1 fault database institute confidence level, net result is derived according to the correlativity between each fault and transition rule, and obtain the travel path of flue gas turbine expander fault, by analyzing the M after each reasoning
avalue and TL obtain the object information of flue gas turbine expander fault propagation forward reasoning; as shown in table 3 below; give when supposing that flue gas turbine expander all faults all occur in table 3; the travel path of each fault of flue gas turbine expander; table 4 lists the change adding fault database institute confidence value before and after protective seam; contrast finds, considers that the fault database institute confidence value after safety device (i.e. protective seam storehouse institute) obviously reduces and more realistic operating mode.The confidence level that field staff can occur according to each fault and travel path carry out emphasis monitoring to the position that may break down, rational maintenance schedule.
Table 3
Table 4
Interpretation of result by inference, draw the confidence level of each root primordium fault causing flue gas turbine expander failure of removal, and it is sorted from big to small, can draw and cause the most probable reason of flue gas turbine expander fault to be due to the bad p6 of shaft assignment or Installation and Debugging are improper, cigarette machine discharge pipe swelling blockage p7 causes cigarette arbor to vibrate excessive p1; The too high p10 of lubricating oil temperature causes the too high p2 of cigarette machine bearing temperature; Stator blade serious scale p16 causes the excessive p3 of cigarette arbor displacement; Further initiation flue gas turbine expander fault.
By comparing the confidence level size of each root primordium fault causing flue gas turbine expander fault, determine the sequencing investigated when there is flue gas turbine expander fault, to remove smoke engine breakdown from root in time, prevent from the rear failure cause of current failure maintenance from still there is the fault caused repeatedly to occur, a working time of increasing device.
Based on above-mentioned diagnostic result, the method for identical fault data end user artificial neural networks is carried out fault diagnosis, diagnostic result contrast is as shown in table 5 below:
Table 5
| Method | Diagnostic result | Evaluation of result |
| Fuzzy fault Petri network | As table 3 | Correctly |
| Artificial neural network | Cannot diagnose, need more sample training | Cannot diagnose |
As can be seen from above-mentioned table 5; adopt this fuzzy fault Petri network method for diagnosing faults can diagnose out some faults cannot diagnosed by existing method exactly; improve the success ratio of Main air compressor fault diagnosis, absolutely prove the feasibility of Main air compressor multifactor relevant failure root degree of depth diagnostic analysis method for Main air compressor fault diagnosis considering protective seam intervention.
Based on same inventive concept, additionally provide a kind of Main air compressor fault rootstock degree of depth diagnostic device in the embodiment of the present invention, as described in the following examples.The principle of dealing with problems due to Main air compressor fault rootstock degree of depth diagnostic device is similar to Main air compressor fault rootstock degree of depth diagnostic method, therefore the enforcement of Main air compressor fault rootstock degree of depth diagnostic device see the enforcement of Main air compressor fault rootstock degree of depth diagnostic method, can repeat part and repeats no more.Following used, term " unit " or " module " can realize the software of predetermined function and/or the combination of hardware.Although the device described by following examples preferably realizes with software, hardware, or the realization of the combination of software and hardware also may and conceived.Fig. 3 is a kind of structured flowchart of the Main air compressor fault rootstock degree of depth diagnostic device of the embodiment of the present invention, as shown in Figure 3, comprise: fuzzy fault Petri network figure sets up unit 301, data capture unit 302, travel path search unit 303 and root primordium determining unit 304, is described below to this structure.
Fuzzy fault Petri network figure sets up unit 301, for according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
Data capture unit 302, for according to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree transition set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
Travel path searches unit 303, for according to described each fault database the set of initial transition confidence level, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Root primordium determining unit 304, for the starting point fault database of travel path that will determine as the root primordium of Main air compressor device fails.
In one embodiment, fuzzy fault Petri network figure sets up unit 301 and comprises: tuple determination module, and set up the tuple needed for described fuzzy fault Petri network figure for determining, wherein, described tuple comprises: P is gathered in P that fault database is gathered, protective seam storehouse
iPL, transition rule set T, fault database event of failure weight vector set w, the threshold value set λ of transition rule, transition rule confidence level set μ, change after TL that fault database is gathered, each fault database initially change confidence level set M
0, each protective seam storehouse crash rate set L and the regular collection R of fuzzy production; Net figure sets up module, for propagating with the rule of the fuzzy production in the regular collection of described fuzzy production, in conjunction with other tuple information in described tuple, set up represent Main air compressor system each fault database between the described fuzzy fault Petri network figure of mutual relationship.
In one embodiment, travel path is searched unit 303 and is comprised: acquisition module, for obtaining initial failure P that storehouse is gathered; First sets up module, for set up with described initial failure storehouse gather transition rule set T corresponding to P; Second sets up module, for set up triggered change after TL that fault database is gathered; Enable transition rule determination module, for determine fault database in described Main air compressor system gather relation in P and transition rule set T between each transition rule, determine current enable transition rule; Transition trigger module, for triggering all current enable transition rules, tries to achieve transition firing sequence; Ask for module, for once changing confidence level set M according to before each fault database institute
a-1, ask for current each fault database institute confidence level set M
a, wherein, occur causing rear collection fault database each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, wherein, a represents transition number of times; Travel path searches module, for the fault database set that obtains after transition is substituted in described fuzzy fault Petri network figure as parameter, find out the maximum fault database of each later reliability travel path.
In one embodiment, enable transition rule determination module specifically for by transition before collect fault database institute to the actual confidence value of transition rule be not less than transition triggering threshold value, and this transition rule is regular before not belonging to the change having triggered transition rule set, then this transition rule is defined as current enable transition rule.
In one embodiment; ask for module specifically for determining whether matcoveredn intervention effect; and when defining protective seam intervention effect, using each protective seam storehouse crash rate set in each protective seam storehouse crash rate as according to one of ask for the confidence level set of current each fault database institute.
In another embodiment, additionally provide a kind of software, this software is for performing the technical scheme described in above-described embodiment and preferred implementation.
In another embodiment, additionally provide a kind of storage medium, store above-mentioned software in this storage medium, this storage medium includes but not limited to: CD, floppy disk, hard disk, scratch pad memory etc.
From above description, can find out, the embodiment of the present invention achieves following technique effect: by each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure, and according to expertise data and on-the-spot actual count data, define each fault database initial trusted degree transition set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, thus it can be used as parameter to substitute into above-mentioned fuzzy fault Petri network figure, can effectively find out the maximum fault database of each later reliability travel path, to determine the root primordium for Main air compressor device fails, thus solve the technical matters that cannot detect the root primordium of Main air compressor device fails in prior art accurately and effectively, reach the technique effect effectively improving accuracy in detection and detection efficiency.Further, because introduce protective seam storehouse institute, make testing result more accurate.
Obviously, those skilled in the art should be understood that, each module of the above-mentioned embodiment of the present invention or each step can realize with general calculation element, they can concentrate on single calculation element, or be distributed on network that multiple calculation element forms, alternatively, they can realize with the executable program code of calculation element, thus, they can be stored and be performed by calculation element in the storage device, and in some cases, step shown or described by can performing with the order be different from herein, or they are made into each integrated circuit modules respectively, or the multiple module in them or step are made into single integrated circuit module to realize.Like this, the embodiment of the present invention is not restricted to any specific hardware and software combination.
The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, for a person skilled in the art, the embodiment of the present invention can have various modifications and variations.Within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. a Main air compressor fault rootstock degree of depth diagnostic method, is characterized in that, comprising:
According to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
According to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
According to described each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Using the starting point fault database in the direction of fault propagation determined as the root primordium of Main air compressor device fails.
2. the method for claim 1, is characterized in that, according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure, comprising:
Determine to set up the tuple needed for described fuzzy fault Petri network figure, wherein, described tuple comprises: P is gathered in P that fault database is gathered, protective seam storehouse
iPL, transition rule set T, fault database event of failure weight vector set w, the threshold value set λ of transition rule, transition rule confidence level set μ, change after TL that fault database is gathered, each fault database initially change confidence level set M
0, each protective seam storehouse crash rate set L and the regular collection R of fuzzy production;
Propagate with the rule of the fuzzy production in the regular collection of described fuzzy production, in conjunction with other tuple information in described tuple, set up represent Main air compressor system each fault database between the described fuzzy fault Petri network figure of mutual relationship.
3. method as claimed in claim 2; it is characterized in that; according to described each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule; find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation, comprising:
Obtain initial failure P that storehouse is gathered;
Set up with described initial failure storehouse gather transition rule set T corresponding to P;
Set up triggered change after TL that fault database is gathered;
Determine fault database in described Main air compressor system gather relation in P and transition rule set T between each transition rule, determine current enable transition rule;
Trigger all current enable transition rules, try to achieve transition firing sequence;
Confidence level set M is once changed according to before each fault database institute
a-1, ask for current each fault database institute confidence level set M
a, wherein, occur causing rear collection fault database each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, wherein, a represents transition number of times;
The fault database set that obtains after transition is substituted in described fuzzy fault Petri network figure as parameter, find out the maximum fault database of each later reliability as the direction of fault propagation.
4. method as claimed in claim 3, is characterized in that, determine current enable transition rule, comprising:
By transition before collect fault database institute to the actual confidence value of transition rule be not less than transition triggering threshold value, and this transition rule do not belong to triggered transition rule set then by this transition rule, be defined as current enable transition rule.
5. method as claimed in claim 3, is characterized in that, ask for current each fault database institute confidence level set M
a, comprising:
Determine whether matcoveredn intervention effect;
When matcoveredn intervention effect, using each protective seam storehouse crash rate set in each protective seam storehouse crash rate as according to one of ask for the confidence level set of current each fault database institute.
6. a Main air compressor fault rootstock degree of depth diagnostic device, is characterized in that, comprising:
Fuzzy fault Petri network figure sets up unit, for according to each fault database of Main air compressor system between mutual relationship, set up fuzzy fault Petri network figure;
Data capture unit, for according to expertise data and on-the-spot actual count data, determine each fault database initial trusted degree set, each fault database the set of event of failure weights, each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule;
Travel path searches unit, for according to described each fault database initial trusted degree set, each fault database the set of event of failure weights and each protective seam storehouse crash rate set, the threshold value set of each transition rule and the confidence level set of transition rule, find out from described fuzzy fault Petri network figure the maximum fault database of each later reliability as the direction of fault propagation;
Root primordium determining unit, for the starting point fault database of travel path that will determine as the root primordium of Main air compressor device fails.
7. device as claimed in claim 6, it is characterized in that, described fuzzy fault Petri network figure sets up unit and comprises:
Tuple determination module, set up the tuple needed for described fuzzy fault Petri network figure for determining, wherein, described tuple comprises: P is gathered in P that fault database is gathered, protective seam storehouse
iPL, transition rule set T, fault database event of failure weight vector set w, the threshold value set λ of transition rule, transition rule confidence level set μ, change after TL that fault database is gathered, each fault database initially change confidence level set M
0, each protective seam storehouse crash rate set L and the regular collection R of fuzzy production;
Net figure sets up module, for propagating with the rule of the fuzzy production in the regular collection of described fuzzy production, in conjunction with other tuple information in described tuple, set up represent Main air compressor system each fault database between the described fuzzy fault Petri network figure of mutual relationship.
8. device as claimed in claim 7, it is characterized in that, described travel path is searched unit and is comprised:
Acquisition module, for obtaining initial failure P that storehouse is gathered;
First sets up module, for set up with described initial failure storehouse gather transition rule set T corresponding to P;
Second sets up module, for set up triggered change after TL that fault database is gathered;
Enable transition rule determination module, for determine fault database in described Main air compressor system gather relation in P and transition rule set T between each transition rule, determine current enable transition rule;
Transition trigger module, for triggering all current enable transition rules, tries to achieve transition firing sequence;
Ask for module, for once changing confidence level set M according to before each fault database institute
a-1, ask for current each fault database institute confidence level set M
a, wherein, occur causing rear collection fault database each front collection fault database institute in confidence level maximal value as transition generation after rear collection fault database confidence value, wherein, a represents transition number of times;
Travel path searches module, for the fault database set that obtains after transition is substituted in described fuzzy fault Petri network figure as parameter, find out the maximum fault database of each later reliability as the direction of fault propagation.
9. device as claimed in claim 8, it is characterized in that, described enable transition rule determination module will be specifically for collecting fault database institute and to be not less than the actual confidence value of transition rule the threshold value of transition triggering before transition, and this transition rule does not belong to and triggered transition rule set and then this transition rule is defined as current enable transition rule.
10. device as claimed in claim 8; it is characterized in that; described module of asking for is specifically for determining whether matcoveredn intervention effect; and when defining protective seam intervention effect, using each protective seam storehouse crash rate set in each protective seam storehouse crash rate as according to one of ask for the confidence level set of current each fault database institute.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410771549.0A CN104571084B (en) | 2014-12-12 | 2014-12-12 | Main air compressor fault rootstock depth diagnostic method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410771549.0A CN104571084B (en) | 2014-12-12 | 2014-12-12 | Main air compressor fault rootstock depth diagnostic method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104571084A true CN104571084A (en) | 2015-04-29 |
| CN104571084B CN104571084B (en) | 2017-07-14 |
Family
ID=53087410
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410771549.0A Active CN104571084B (en) | 2014-12-12 | 2014-12-12 | Main air compressor fault rootstock depth diagnostic method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104571084B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005644A (en) * | 2015-06-29 | 2015-10-28 | 山东科技大学 | Method for detecting fault of three-phase asynchronous motor |
| CN107203202A (en) * | 2017-07-03 | 2017-09-26 | 贵州大学 | Systems reliability analysis and method for diagnosing faults based on Fuzzy Petri Net |
| CN110672198A (en) * | 2019-08-26 | 2020-01-10 | 华电电力科学研究院有限公司 | Boiler flue gas and air system vibration fault diagnosis method |
| CN110989561A (en) * | 2019-12-26 | 2020-04-10 | 中国航空工业集团公司沈阳飞机设计研究所 | Method for constructing fault propagation model and method for determining fault propagation path |
| CN111998690A (en) * | 2020-08-24 | 2020-11-27 | 中国测试技术研究院 | Fault alarm method for smoke exhaust fan of low-temperature kiln |
| CN112147974A (en) * | 2020-07-22 | 2020-12-29 | 中国石油大学(华东) | Alarm root cause diagnosis method based on chemical process knowledge automation |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1049206A (en) * | 1996-07-31 | 1998-02-20 | Kobe Steel Ltd | Sequence control generating device |
| WO2006031749A2 (en) * | 2004-09-10 | 2006-03-23 | Exxonmobil Research And Engineering Company | Application of abnormal event detection technology to fluidized catalytic cracking unit |
| CN104156627A (en) * | 2014-08-29 | 2014-11-19 | 中国石油大学(北京) | Fault diagnosis method and device based on device working condition |
-
2014
- 2014-12-12 CN CN201410771549.0A patent/CN104571084B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH1049206A (en) * | 1996-07-31 | 1998-02-20 | Kobe Steel Ltd | Sequence control generating device |
| WO2006031749A2 (en) * | 2004-09-10 | 2006-03-23 | Exxonmobil Research And Engineering Company | Application of abnormal event detection technology to fluidized catalytic cracking unit |
| CN104156627A (en) * | 2014-08-29 | 2014-11-19 | 中国石油大学(北京) | Fault diagnosis method and device based on device working condition |
Non-Patent Citations (1)
| Title |
|---|
| JINQIU HU ETC.: "Root Cause Reasoning Based on Fuzzy Fault Petri Net For Petroleum Engineering Systems", 《2014 11TH INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS AND KNOWLEDGE DISCOVERY》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005644A (en) * | 2015-06-29 | 2015-10-28 | 山东科技大学 | Method for detecting fault of three-phase asynchronous motor |
| CN105005644B (en) * | 2015-06-29 | 2019-03-01 | 山东科技大学 | A method of detection threephase asynchronous failure |
| CN107203202A (en) * | 2017-07-03 | 2017-09-26 | 贵州大学 | Systems reliability analysis and method for diagnosing faults based on Fuzzy Petri Net |
| CN110672198A (en) * | 2019-08-26 | 2020-01-10 | 华电电力科学研究院有限公司 | Boiler flue gas and air system vibration fault diagnosis method |
| CN110672198B (en) * | 2019-08-26 | 2021-08-17 | 华电电力科学研究院有限公司 | Boiler flue gas and air system vibration fault diagnosis method |
| CN110989561A (en) * | 2019-12-26 | 2020-04-10 | 中国航空工业集团公司沈阳飞机设计研究所 | Method for constructing fault propagation model and method for determining fault propagation path |
| CN112147974A (en) * | 2020-07-22 | 2020-12-29 | 中国石油大学(华东) | Alarm root cause diagnosis method based on chemical process knowledge automation |
| CN112147974B (en) * | 2020-07-22 | 2022-07-12 | 中国石油大学(华东) | Alarm root cause diagnosis method based on chemical process knowledge automation |
| CN111998690A (en) * | 2020-08-24 | 2020-11-27 | 中国测试技术研究院 | Fault alarm method for smoke exhaust fan of low-temperature kiln |
| CN111998690B (en) * | 2020-08-24 | 2021-05-25 | 中国测试技术研究院 | Fault alarm method for smoke exhaust fan of low-temperature kiln |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104571084B (en) | 2017-07-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104571084A (en) | Further diagnosis method and device for root causes of failure of main fan set | |
| Zaher et al. | A multi-agent fault detection system for wind turbine defect recognition and diagnosis | |
| CN102819239B (en) | Intelligent fault diagnosis method of numerical control machine tool | |
| CN106054104A (en) | Intelligent ammeter fault real time prediction method based on decision-making tree | |
| Dao | On Wilcoxon rank sum test for condition monitoring and fault detection of wind turbines | |
| CN106124982A (en) | Automatic expert's resultant fault diagnostic system of a kind of Wind turbines and diagnostic method | |
| CN103150473A (en) | Method and device for monitoring and diagnosing generating efficiency of wind turbine generator in real-time manner | |
| CN103278328B (en) | Method for diagnosing failure of hydraulic variable-pitch system of wind turbine generator based on fuzzy Petri net | |
| CN101783749B (en) | Network fault positioning method and device | |
| CN110119518B (en) | Method for diagnosing engine fault cause by adopting neural network model | |
| CN104765965A (en) | GIS fault diagnosis and reliability analysis method based on fuzzy Petri | |
| CN109815441B (en) | Method for diagnosing and predicting engine faults by adopting Bayesian network model | |
| CN109492790A (en) | Wind turbines health control method based on neural network and data mining | |
| Ahsan et al. | Reliability analysis of gas turbine engine by means of bathtub‐shaped failure rate distribution | |
| CN102252843A (en) | Assessment method for rolling bearing performance variation | |
| CN116862081B (en) | Operation and maintenance method and system for pollution treatment equipment | |
| CN112072642B (en) | Method for estimating abnormal operation state of power system | |
| CN116670608A (en) | Hybrid ensemble method for predictive modeling of Internet of things | |
| Du et al. | A SCADA data based anomaly detection method for wind turbines | |
| CN113395182B (en) | Intelligent network equipment management system and method with fault prediction | |
| Strasser et al. | Diagnostic alarm sequence maturation in timed failure propagation graphs | |
| CN112785010B (en) | Multi-component dynamic maintenance strategy optimization method for rail transit train system | |
| CN114320773B (en) | Wind turbine generator system fault early warning method based on power curve analysis and neural network | |
| CN107727392A (en) | State index evaluation and optimization method based on signal detection and ROC analysis | |
| Wang et al. | A method of searching fault propagation paths in mechatronic systems based on MPPS model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |