CN105159286B - A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system - Google Patents
A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system Download PDFInfo
- Publication number
- CN105159286B CN105159286B CN201510608592.XA CN201510608592A CN105159286B CN 105159286 B CN105159286 B CN 105159286B CN 201510608592 A CN201510608592 A CN 201510608592A CN 105159286 B CN105159286 B CN 105159286B
- Authority
- CN
- China
- Prior art keywords
- reasoning
- confidence level
- telemetry
- knowledge
- result
- 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
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
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
- G05B23/0227—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions
- G05B23/0229—Qualitative history assessment, whereby the type of data acted upon, e.g. waveforms, images or patterns, is not relevant, e.g. rule based assessment; if-then decisions knowledge based, e.g. expert systems; genetic algorithms
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Test And Diagnosis Of Digital Computers (AREA)
- Alarm Systems (AREA)
Abstract
The present invention discloses a kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system, wherein：Knowledge editor inputs compiled good alarm diagnosis knowledge；The original telemetry of data buffer zone buffering input；Data field stores from the telemetry during extraction reasoning and decision device progress logic matching operation of data buffer zone or telemetry command and to it；Formula area loads compiled alarm diagnosis knowledge from knowledge editor, and in formula area, each alarm diagnosis knowledge is referred to as rule；Alarm diagnosis knowledge in original telemetry in data field and formula area is carried out logic matching operation and obtains diagnostic result by reasoning and decision device；The diagnostic result that selection needs to export is exported to blackboard；The diagnostic result that blackboard storage reasoning and decision device is obtained by logic matching operation；The diagnostic result of result buffer buffering reasoning and decision device selection output is simultaneously sent to client, and confirmation is replied after user checks diagnostic result by the client.
Description
Technical field
The invention belongs to space flight fault diagnosis technology field, more particularly to a kind of in-orbit abnormal alarm of spacecraft to examine with failure
Disconnected system.
Background technology
The core of inference machine subsystem is to carry out reasoning from logic computing according to expertise and the measuring and control data of input, finally
The abnormality of spacecraft is released, its essence is realize a virtual machine with logical reasoning ability.The expert system of early stage
It is main to use certainty reasoning algorithm, the being to determine property of evidence that this method represents.Such as it is not light by taking ground shadow state as an example
According to area be exactly shadow zone, without ambiguity.Its reasoning process is based on mathematical logic, and reasoning process is tight, conclusion
And it is accurate, otherwise set up, otherwise it is invalid.But the evidence used during the diagnosis of reality is not completely accurate
, some information are not perfect enough, and some have uncertainty.Such as in shadow state, the description for penumbra region is just relatively more tired
Difficulty, penumbra since when shadow at last, be difficult description to this using Accurate Reasoning.
The content of the invention
To solve the above problems, the present invention provides a kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system.
The in-orbit abnormal alarm of spacecraft and fault diagnosis system of the present invention, it includes：Knowledge editor, data buffering
Area, reasoning and decision device, data field, blackboard, formula area and result buffer；
Knowledge editor, for inputting compiled alarm diagnosis knowledge；
Data buffer zone, for buffering the original telemetry of input, original telemetry includes telemetry and remote measurement
Instruction；
Data field, for extracting the telemetry or distant when reasoning and decision device carries out logic matching operation from data buffer zone
Survey and instruct and it is stored；
Formula area, for loading compiled alarm diagnosis knowledge from knowledge editor, each reported in formula area
Alert diagnostic knowledge is referred to as rule；
Reasoning and decision device, the alarm diagnosis knowledge in the original telemetry in data field and formula area is subjected to logic
With computing, diagnostic result is obtained；And the diagnostic result for selecting to need to export is exported to blackboard；
Blackboard, the diagnostic result obtained for storing reasoning and decision device by logic matching operation；
Result buffer, for buffering the diagnostic result of reasoning and decision device selection output and sending to client, treat user
Confirmation is replied after checking diagnostic result by the client.
Further, reasoning and decision device includes：
Reliability assessment module, the assessment for the parameter confidence level that taken remote measurement to the original telemetry of input, is obtained with credible
The telemetry of degree；Wherein, the initial value table information and the original telemetry that telecommand, inference machine are safeguarded are together as space flight
The fact that device fault diagnosis reasoning；
Evidence derivation module, the initial value table information and the band confidence level safeguarded according to the telecommand on ground, inference machine
Telemetry carry out logical operation and obtain evidence, so-called evidence is exactly according to the true status information drawn with factual knowledge
With warning message；
Fault diagnosis module, fault diagnosis is carried out according to the evidence combination alarm diagnosis knowledge, releases diagnostic result；
As a result output module, according to alarm diagnosis knowledge-chosen diagnostic result and export.
Wherein, described reliability assessment module includes：
Rule statistic unit, remote measurement statistical law is obtained by carrying out statistics to the changing rule of telemetry parameter；
Confidence level obtaining unit, by the remote measurement statistical law of new telemetry parameter and the progress of the remote measurement statistical law of history
Match somebody with somebody, and combine the confidence level that confidence factor obtains remote measurement.
Wherein, the evidence derivation module includes：
The derivation that transfiniting based on degree of membership judges and the confidence level pass-algorithm two ways based on fuzzy mathematics is carried out,
Wherein, the confidence level pass-algorithm based on fuzzy mathematics includes：Single true confidence level pass-algorithm and combination are true to be added
Right reliabiliL pass-algorithm.
Wherein, the fault diagnosis module includes：Forward reasoning, backward reasoning, the framework rule-based reasoning based on possibility.
Wherein, the output principle as a result in output module is：Confidence level more height output speed is faster；The result of low confidence level
If the long time is maintained to be also required to export；Influence of the time nearer data to output result is bigger；Confidence level is 0.9
When output time be 1 minute；Output time is 1 hour when confidence level is 0.6；Confidence level be more than 0.5 support result output, can
Reliability is less than not exporting for 0.5 support result.
Further, each rule is provided with activation control list in formula area, and the formula area is first judged when making inferences
In regular activation control list whether set up, establishment then enables the rule, otherwise will not be made inferences from the rule
Computing.
Wherein, the fault diagnosis module includes：Real-time diagnosis unit, early warning diagnosis unit, play back diagnosis unit, checking
Diagnosis unit and spread function unit.
Beneficial effect：
The present invention can be inferred that whether spacecraft occurs exception, can automatically search for, is matched accordingly when an exception occurs
Expertise, and provide alarm failure positioning and failure aid in treatment decision information.And the present invention can be realized to history
Telemetry reexamines, the alarm of ex-post analysis Spacecraft anomaly, diagnosis positioning and failure aid in treatment decision-making.
The system is adapted to remote measurement analog quantity and the class parameter of quantity of state two, can according to history telemetry data conversion rule,
The reliability coefficient and confidence level of telemetry are calculated, the degree of accuracy of Analysis on confidence should reach more than 95%.
Brief description of the drawings
Fig. 1 is the in-orbit abnormal alarm of spacecraft and fault diagnosis system schematic diagram of the present invention；
Fig. 2 is the real-time diagnosis cell schematics of the present invention；
Fig. 3 is the playback diagnosis unit schematic diagram of the present invention；
Fig. 4 is the early warning diagnosis unit schematic diagram of the present invention；
Fig. 5 is the checking diagnosis unit schematic diagram of the present invention；
The transfiniting based on degree of membership that Fig. 6 is the present invention judges the schematic diagram of embodiment one.
Embodiment
Substantial amounts of fuzzy concept when spacecraft fault diagnosis knowledge description be present, such as be exactly a ratio for transfiniting
Relatively fuzzy concept, it is to transfinite that numerical value, which is how many, how many seriously to transfinite, and can not accurately be defined, therefore go back in the present system
Complex reasoning is carried out to these fuzzy concepts using the subordination method in fuzzy reasoning.Except entering when mathematical operation
Row numerical operation also needs to the confidence level with fuzzy mathematics evaluation.Therefore also need to combine fuzzy push away on reasoning algorithm
Some contents in reason.Finally when diagnostic result is exported, when is exported, it is necessary to be built according to user behavior analysis
Vertical empirical model, ultimately forms spacecraft fault diagnosis mathematical modeling.
Specific implementation is as follows：
As shown in figure 1, the in-orbit abnormal alarm of spacecraft and fault diagnosis system of the present invention, it includes：Knowledge editor,
Data buffer zone, reasoning and decision device, data field, blackboard, formula area and result buffer；
Knowledge editor, for inputting compiled alarm diagnosis knowledge；
Data buffer zone, for buffering the original telemetry of input, original telemetry includes telemetry and remote measurement
Instruction；
Data field, for extracting the telemetry or distant when reasoning and decision device carries out logic matching operation from data buffer zone
Survey and instruct and it is stored；
Formula area, for loading compiled alarm diagnosis knowledge from knowledge editor, each reported in formula area
Alert diagnostic knowledge is referred to as rule；
Reasoning and decision device, the alarm diagnosis knowledge in the original telemetry in data field and formula area is subjected to logic
With computing, diagnostic result is obtained；And the diagnostic result for selecting to need to export is exported to blackboard；
Blackboard, the diagnostic result obtained for storing reasoning and decision device by logic matching operation；
Result buffer, for buffering the diagnostic result of reasoning and decision device selection output and sending to client, treat user
Confirmation is replied after checking diagnostic result by the client.
Further, reasoning and decision device includes：
Reliability assessment module, the assessment for the parameter confidence level that taken remote measurement to the original telemetry of input, is obtained with credible
The telemetry of degree；Wherein, the initial value table information and the original telemetry that telecommand, inference machine are safeguarded are together as space flight
The fact that device fault diagnosis reasoning；The main purpose of telemetry parameter reliability assessment is to assess to observe true credibility,
By that can have certain error code after telemetry ground receiver, Analysis on confidence is exactly the credible journey for judging these data
Degree, each each data stamp confidence level label, represent the degree of support to evidence, are sent into inference machine and make inferences computing.It is credible
The theoretical foundation that degree is assessed is probability theory.By being counted to the changing rule of telemetry parameter, if receiving new remote measurement
After data, judge that the Compound Degree of remote measurement statistical law of the new telemetry changing rule with counting is matched, if met
It is then high confidence level telemetry to compare high, is otherwise low confidence level.Telemetry due to error code largely
The result of reasoning is influenceed, especially in the near-earth satellite time out of the station, the bit error rate of data is very high, may flood normal data,
This when, inference conclusion was substantially what can not be differentiated, it is therefore desirable to analyzes telemetry and show that believable one of data are
Number.This coefficient participates in reasoning as the confidence factor of factural information, and conclusion can be finally released according to this confidence level
Confidence level.
Evidence derivation module, the initial value table information and the band confidence level safeguarded according to the telecommand on ground, inference machine
Telemetry carry out logical operation and obtain evidence, so-called evidence is exactly according to the true status information drawn with factual knowledge
With warning message；Generation evidence is exactly according to the true process that evidence is formed with factual knowledge, is exactly foundation in the present system
Telemetry and the alarm knowledge of telemetry parameter form the process of warning message.The reasoning computing of generation evidence uses fuzzy reasoning
Method, main is exactly to use subordination method, and the transmission method of confidence level.Introduce separately below：
Fault diagnosis module, fault diagnosis is carried out according to the evidence combination alarm diagnosis knowledge, releases diagnostic result；
As a result output module, according to alarm diagnosis knowledge-chosen diagnostic result and export.For how to determine the reasoning results
Output, the behavior monitored by assayer to spacecraft, it is assumed that there is following hypothesis：
The confidence level computational mathematics model of output result is applied not only to the output of confidence level result, be additionally operable to process monitoring with
The shooting condition inspection of troubleshooting monitoring.Excite and the output condition of process monitoring have certain similitude, are all satisfactions one
The condition of fixing time is exported or carried out next step action.
Further, described reliability assessment module includes：
Rule statistic unit, remote measurement statistical law is obtained by carrying out statistics to the changing rule of telemetry parameter；
Confidence level obtaining unit, by the remote measurement statistical law of new telemetry parameter and the progress of the remote measurement statistical law of history
Match somebody with somebody, and combine the confidence level that confidence factor obtains remote measurement.
Wherein, the evidence derivation module includes：
The derivation that transfiniting based on degree of membership judges and the confidence level pass-algorithm two ways based on fuzzy mathematics is carried out,
Wherein, the confidence level pass-algorithm based on fuzzy mathematics includes：Single true confidence level pass-algorithm and combination are true to be added
Right reliabiliL pass-algorithm.
Transfiniting based on degree of membership judges embodiment one
Example 1：Area of illumination busbar voltage reduces (under-voltage)
42V ± 0.5V is should be in area of illumination busbar voltage normal value.The reason for causing voltage to reduce has：
(1) parallel regulator failure causes shunting value excessive, and solar battery array power output is split, busbar voltage control
Failure, it is impossible to be load supplying.Check the output of bus error signal：Northern VN4 (southern VN12) >=6V.Whether -12V power supplies open a way
(remote measurement code name).
(2) moon shade is entered.(normal phenomenon, being calculated according to CALCULATING PREDICTION)
(3) load current excessively stream on bus.It can not recover before unshorting factor normal
(4) solar array is reversed, stalled.Recover windsurfing normal condition.
According to the discussion research with user and actual conditions, handled for the two parts that are anomaly divided into of spacecraft,
Alarm first, that is, for telemetry parameter transfiniting under numerous conditions, spacecraft state transition is detected, as report
Alert result is exported；Then it is exactly the diagnosis of failure, failure is positioned in detail, analyzes the reason for failure occurs, finally
To out of order solution.
The fault alarm just reduced below according to area of illumination busbar voltage diagnoses, and is analyzed, and sorts out alarm knowledge (thing
Real sex knowledge) and fault diagnosis knowledge, the description of process is then made inferences again.
Fact knowledge
Ra1：If being less than 41.5V in area of illumination busbar voltage, area of illumination busbar voltage reduces.
Ra2：Reduced in area of illumination busbar voltage, if being in moon shadow state, normally.
Ra3：Reduced in area of illumination busbar voltage, if solar array reverses or stalling, normally.The alarm of solar array
Rule can quote windsurfing and reverse or stall.
Diagnostic knowledge
Rd1：If area of illumination busbar voltage reduces, it is excessive to judge whether current divider shunts, is current divider if excessive
Failure causes busbar voltage to reduce；Otherwise busbar voltage is caused to reduce for unknown cause.
Rd2：If area of illumination busbar voltage reduces, judge whether bus load current is excessive, be short circuit if excessive
Voltage is caused to reduce；Otherwise busbar voltage is caused to reduce for unknown cause.
Then regard as transfiniting less than 41.5V for area of illumination busbar voltage in alarm rule Ra1, for this rule generally
Judgement transfinite less than 41.5, do not transfinited more than 41.5, realize also fairly simple, but had a problem that, if bus
Voltage is whether 41.50001 be also to transfinite.For this problem it may be said that transfiniting, alternatively transfiniting (sampling error).Cause
The reason for this problem is a Ra1 inherently fuzzy concept, if using accurate description method, when parameter is in critical zone
When would become hard to describe, the selection of critical value has large effect to result.There is ratio in fuzzy mathematics in this case
Preferably solve method, judgement of transfiniting exactly is described using membership function.As shown in Figure 6 be exactly busbar voltage be less than 41.5 surpass
Limit the degree of membership curve judged.Curve expression in figure 6 judges the degree of membership curve of the super lower limit of busbar voltage, and abscissa is mother
Line voltage value, ordinate are the degree that transfinites.
From the graph it can be seen that the degree of transfiniting is 80% when busbar voltage is 41.5, if parameter be 41.4 when if
The degree that transfinites is 95%, if representing that the degree of transfiniting is 4% when parameter is 41.6.Transfinited if parameter is 41.50001
Degree is close to 80%, and (system integrates to the degree of transfiniting, and reaches and excites to a certain degree if the duration is longer
Subsequent action), warning output can be equally excited, or excite follow-up reasoning process.
If from physical significance exactly ought the degree of transfiniting reach 80% and be super lower limit if maintaining the regular hour；Such as
It is also then super lower limit that fruit, which is less than 80% and maintains the long time,；The time that if degree that transfinites is higher than 80% and maintenance is shorter
Also it is super lower limit.This physical significance also complies with the mode of thinking that expert's observation is transfinited, that is, comparison of transfiniting is high, then can recognize immediately
Being set to transfinites；If it is not serious to transfinite, but continues up, then it is also assumed that being super lower limit.
For the analysis of case above, it will be seen that using the judgement of transfiniting based on degree of membership, can be very good to solve
The critical value fuzzy problem of certainly simple overload alarm
Confidence level pass-algorithm embodiment based on fuzzy mathematics
1) single true confidence level pass-algorithm：
If the fact that support conclusion only has one, and known true E confidence level CF (E) and factual knowledge
IF E THEN H
Confidence level CF (H, E), then conclusion H confidence level be exactly：
CF (H)=CF (E) * CF (H, E)
2) the true weighting confidence level pass-algorithm of combination
If it is known that the fact has multiple one true E of combination of formation
E=E1 ∧ E2 ∧ E3 ∧ ... ∧ En,
And known each true confidence level CF (Ei), and each true weight coefficient P in this combination is true
, and factual knowledge (Ei)
IF E THEN H
Confidence level CF (H, E), then conclusion H confidence level be exactly：
Further, the fault diagnosis module includes：Forward reasoning, backward reasoning, the framework rule based on possibility
Reasoning.
Forward reasoning：Mainly carried out in the present system using forward reasoning mode, that is, from the fact, by logic
Reasoning, finally draw inference conclusion.It is measuring and control data for the main foundation of spacecraft fault diagnosis, when receiving new observing and controlling
When data, start reasoning.
Backward reasoning is that conclusion solves process, a conclusion is assumed first that when reasoning, then searches this knot
By whether setting up, it is considered that this conclusion is to set up if setting up.In spacecraft fault diagnosis, for state transition
Reasoning using reverse manner carry out.The shooting condition of reasoning is that saltus step occurs for state, and this when, system was assumed to state
Normal variation (automatic control or remote control), then searches and meets whether the condition of normal variation meets, is exactly normal become if meeting
Change, be otherwise anomalous variation.For backward reasoning, it is desirable to which conclusion is easy to it is assumed that so reasoning has specific aim, thrust
Can be relatively good.
Framework rule-based reasoning based on possibility：In spacecraft fault diagnosis language, there is a kind of diagnostic knowledge to use frame
Frame mode describes, and is exactly first to write setting failure, in the form of expression of description failure.Such as the excessive shunting failure of current divider,
Its phenomenon is to be reduced in area of illumination busbar voltage, and shunt current is higher.For species knowledge, if phenomenon does not all go out
Existing, then rule will not activate；Once there is a phenomenon to occur, then assume that failure occurs, calculate the confidence level of failure conclusion immediately,
If Reliability ratio is higher, then it is assumed that releases conclusion.Frame inference strategy is determined by the input mode of knowledge.
Wherein, the output principle as a result in output module is：Confidence level more height output speed is faster；The result of low confidence level
If the long time is maintained to be also required to export；Influence of the time nearer data to output result is bigger；Confidence level is 0.9
When output time be 1 minute；Output time is 1 hour when confidence level is 0.6；Confidence level be more than 0.5 support result output, can
Reliability is less than not exporting for 0.5 support result.The empirical equation that diagnostic result output confidence level calculates：
Wherein
OfRepresent the confidence level of diagnostic result output；
Cf(t, h) represents confidence levels of the result diagnostic result h in t；
μtRepresent disturbance degree of the t to output, it is believed that be a time weight value, be a membership function.
μt=et。
Further, each rule is provided with activation control list in formula area, and the formula area is first judged when making inferences
In regular activation control list whether set up, establishment then enables the rule, otherwise will not be made inferences from the rule
Computing.
Further, the fault diagnosis module includes：Real-time diagnosis unit, early warning diagnosis unit, diagnosis unit is played back,
Verify diagnosis unit and spread function unit.
Real-time diagnosis unit：Real time execution pattern is system primary operating mode, realizes the real-time diagnosis report to spacecraft
It is alert, the running of real time execution pattern is described in fig. 2.(1) start：Monitoring and scheduling module starts inference machine, passes through parameter
Real-time mode operation is arranged to, inference machine determines the preserving type of the acquisition modes of data, the reasoning results according to mode parameter.
(2) diagnostic knowledge is loaded：Diagnostic knowledge is loaded from database when inference machine starts, builds various reasoning objects, completion pair
The initialization procedure of inference machine.(3) initial value table is loaded：The current state of spacecraft is recorded in system in real-time mode, is being opened
Load these status informations simultaneously when dynamic, for comparing the abnormal saltus step of spacecraft state.(4) measuring and control data is sent：In reasoning
When machine is run, measuring and control data that real-time reception measuring and control data sending module is sent, mainly telecommand data herein.
(5) measuring and control data is sent：Measuring and control data sending module sends telemetry and is sent to telemetry Analysis on confidence module simultaneously,
For analyzing the confidence level of telemetry, inference machine is sent to by network again after the completion of analysis.(6) measuring and control data is sent：It is distant
Survey data reliability analysis module and the telemetry parameter for stamping confidence level label is sent to inference machine.Inference machine is used with credible
The data of scale label make inferences.(7) history measuring and control data is read：Enter in reasoning process sometimes for history measuring and control data
OK, directly read by satellite integrated database, and cached in systems, for performance when history telemetry is read
Needs, it is necessary to pre-read to data, ensure the speed of reasoning.(8) diagnostic result is preserved：Finished to a frame reasoning
Afterwards, the result of release is analyzed, using these the reasoning results as in diagnostic result write into Databasce.(9) diagnostic result is sent：Reasoning
After machine infers new diagnostic result, alarm diagnosis client is sent to by network, user checks diagnosis knot by browser
Fruit.(10) diagnostic result is confirmed：After user checks diagnostic result, according to the correctness for explaining information judged result, and to diagnosis
As a result confirmed, represent that user has known this result, client will confirm that result is sent to inference machine by network, push away
Reason machine is according to the confidence level of the confirmation results modification the reasoning results of user, the reasoning for next step
Diagnosis unit is played back, measuring and control data carries out diagnosis point during inference machine is to historical data base in data readback pattern
Analysis, monitor the exception once occurred of spacecraft.Inference machine is started by monitoring and scheduling module with playback mode, and user passes through playback
The time end of diagnostic clients end input write-in playback, and start reasoning process.Inference machine is directly from defending in diagnostic mode is played back
Measuring and control data is read in star integrated database and carries out diagnostic analysis, and sends result to alarm diagnosis client.In playback mould
After formula is run, inference machine process automatically exits from running.The running of data readback pattern is described in figure 3.Start, hair
Control command, display reasoning operation progress msg are sent, diagnostic knowledge is loaded, reads playback measuring and control data, diagnostic reasoning, is stored back into
Put result, display playback result.
Early warning diagnosis unit：, it is necessary to there is two inference machine operations when modes of warning is run, one is run in real-time mode,
Another runs in modes of warning.Telemetry prediction module is extrapolated to remote measurement numerical value after receiving telemetry, outside
The spacecraft status information needed when pushing away needs the inference machine from real-time mode operation to obtain.Remote measurement prediction data is sent to
The inference machine of modes of warning operation, inference machine carry out diagnostic analysis according to these prediction data, while receive real-time mode transmission
The diagnostic result come, the result for finally not exiting real-time mode is sent to alarm indication client, with the shape of early warning result
Formula is shown.Modes of warning inference machine and real-time mode operation inference machine all in the form of services continuous service backstage, by adjusting
Spend the running situation of monitoring module monitoring process.The running of modes of warning is described in Fig. 4.Start, load diagnostic knowledge,
Measuring and control data, real-time diagnosis are sent, real-time status result is sent, calculates prediction data, early warning diagnosis, storage early warning diagnosis knot
Fruit, display early warning result.
Verify diagnosis unit：In Verify in System pattern, system verifies that client is read to measuring and control data, then root from system
According to the analog parameter rule of user's input, generation checking data, further according to the transmission rate of real-time telemetry data, with telemetry frame
Form is sent to inference machine.Inference machine is run with Validation Mode, is received checking data and is carried out diagnostic reasoning, finally by diagnostic result
Alarm diagnosis client is sent to show.
Verify in System modular system verifies the clock of client maintenance telemetry, it is desirable to when inference machine is also according to this
Clock makes inferences.After the completion of checking, inference machine process automatically exits from.The operation of data system Validation Mode is described in Figure 5
Process.Start, start checking inference machine, selection checking data, send checking data, checking diagnosis, storage the result, display
The result.
Further, the inference step in reasoning and decision device includes：
Step 0, loading diagnostic rule f1-fm, regular fi is made up of n evidence and a conclusion；
Step 1, the original telemetry x1---xn in n evidence is assessed using " patent " for regular fi, obtained
Obtain telemetry confidence level y1---yn；
Step 2, for i-th of evidence, the evidence for prestoring original telemetry xi judges fuzzy interval ci, based on original
Beginning telemetry xi calculates original telemetry xi using degree of membership algorithm and judges the reliability coefficient in fuzzy interval in evidence
di；
Reliability coefficient di is multiplied with telemetry confidence level yi and obtains Certainty Factor ei；
Step 3, repeat step 2, the calculating of all Certainty Factor e1-en in regular fi is completed；
Step 4, the decision confidence for regular fi is calculated using confidence level pass-algorithm；
Step 5, whether the decision confidence computation rule fi obtained using assessment algorithm combination step 4 exports alarm signal
Breath.
Step 6, repeat step 1 to 5, the reasoning process of strictly all rules is completed.
Certainly, the present invention can also have other various embodiments, ripe in the case of without departing substantially from spirit of the invention and its essence
Know those skilled in the art when can be made according to the present invention it is various it is corresponding change and deformation, but these corresponding change and become
Shape should all belong to the protection domain of appended claims of the invention.
Claims (7)
1. a kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system, it is characterised in that including：Knowledge editor, data are delayed
Rush area, reasoning and decision device, data field, blackboard, formula area and result buffer；
Knowledge editor, for inputting compiled alarm diagnosis knowledge；
Data buffer zone, for buffering the original telemetry of input, original telemetry includes telemetry and telemetry command；
Data field, for referring to from the telemetry during extraction reasoning and decision device progress logic matching operation of data buffer zone or remote measurement
Make and it is stored；
Formula area, for loading compiled alarm diagnosis knowledge from knowledge editor, each alarm and examine in formula area
Disconnected knowledge is referred to as rule；
Reasoning and decision device, the original telemetry in data field is subjected to logic with the alarm diagnosis knowledge in formula area and matches fortune
Calculate, obtain diagnostic result；And the diagnostic result for selecting to need to export is exported to blackboard；
Blackboard, the diagnostic result obtained for storing reasoning and decision device by logic matching operation；
Result buffer, for buffering the diagnostic result of reasoning and decision device selection output and sending to client, treat that user passes through
The client replys confirmation after checking diagnostic result；
The reasoning and decision device includes：
Reliability assessment module, the assessment for the parameter confidence level that taken remote measurement to the original telemetry of input, is obtained with confidence level
Telemetry；Wherein, the initial value table information and the original telemetry that telecommand, inference machine are safeguarded are together as spacecraft event
The fact that hinder diagnostic reasoning；
Evidence derivation module, the initial value table information and described with the distant of confidence level safeguarded according to the telecommand on ground, inference machine
Survey data progress logical operation and obtain evidence, so-called evidence is exactly according to the true status information drawn with factual knowledge and report
Alert information；
Fault diagnosis module, fault diagnosis is carried out according to the evidence combination alarm diagnosis knowledge, releases diagnostic result；
As a result output module, according to alarm diagnosis knowledge-chosen diagnostic result and export.
2. the in-orbit abnormal alarm of spacecraft as claimed in claim 1 and fault diagnosis system, it is characterised in that described is credible
Degree evaluation module includes：
Rule statistic unit, remote measurement statistical law is obtained by carrying out statistics to the changing rule of telemetry parameter；
Confidence level obtaining unit, the remote measurement statistical law of new telemetry parameter is matched with the remote measurement statistical law of history, and
The confidence level of remote measurement is obtained with reference to confidence factor.
3. the in-orbit abnormal alarm of spacecraft as claimed in claim 1 and fault diagnosis system, it is characterised in that the evidence pushes away
Guide module includes：
The derivation that transfiniting based on degree of membership judges and the confidence level pass-algorithm two ways based on fuzzy mathematics is carried out, its
In, the confidence level pass-algorithm based on fuzzy mathematics includes：The weighting of single true confidence level pass-algorithm and the combination fact
Confidence level pass-algorithm.
4. the in-orbit abnormal alarm of spacecraft as claimed in claim 1 and fault diagnosis system, it is characterised in that the failure is examined
Disconnected module includes：Forward reasoning, backward reasoning, the framework rule-based reasoning based on possibility.
5. the in-orbit abnormal alarm of spacecraft as claimed in claim 1 and fault diagnosis system, it is characterised in that result exports mould
Output principle in block is：Confidence level more height output speed is faster；If the result of low confidence level maintains the long time
Need to export；Influence of the time nearer data to output result is bigger；Output time is 1 minute when confidence level is 0.9；It is credible
Spend for 0.6 when output time be 1 hour；Confidence level is more than 0.5 and supports the output of result, confidence level to be less than 0.5 and support result
Do not export.
6. the in-orbit abnormal alarm of spacecraft as claimed in claim 1 and fault diagnosis system, it is characterised in that every in formula area
Individual rule is provided with activation control list, first judges whether the regular activation control list in the formula area is equal when making inferences
Set up, establishment then enables the rule, otherwise will not make inferences computing from the rule.
7. the in-orbit abnormal alarm of spacecraft as claimed in claim 4 and fault diagnosis system, it is characterised in that the failure is examined
Disconnected module includes：Real-time diagnosis unit, early warning diagnosis unit, diagnosis unit is played back, verify diagnosis unit and spread function unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510608592.XA CN105159286B (en) | 2015-09-22 | 2015-09-22 | A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510608592.XA CN105159286B (en) | 2015-09-22 | 2015-09-22 | A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105159286A CN105159286A (en) | 2015-12-16 |
| CN105159286B true CN105159286B (en) | 2017-12-08 |
Family
ID=54800171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510608592.XA Active CN105159286B (en) | 2015-09-22 | 2015-09-22 | A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105159286B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106371030A (en) * | 2016-08-29 | 2017-02-01 | 丹阳亿豪电子科技有限公司 | New energy automobile battery fault diagnosis method based on uncertainty reasoning |
| CN106647695A (en) * | 2016-12-05 | 2017-05-10 | 航天恒星科技有限公司 | Method and system for monitoring running state of on-orbit spacecraft |
| CN110333710B (en) * | 2019-06-28 | 2020-12-18 | 中国空间技术研究院 | System and method for detecting and processing on-orbit fault of spacecraft |
| CN111240966B (en) * | 2020-01-03 | 2023-10-24 | 中国建设银行股份有限公司 | Alarm information processing method and system |
| CN111193474B (en) * | 2020-01-14 | 2020-11-20 | 北京空间飞行器总体设计部 | High-precision autonomous diagnosis method for satellite solar wing output current |
| CN111274543A (en) * | 2020-01-17 | 2020-06-12 | 北京空间飞行器总体设计部 | Spacecraft system anomaly detection method based on high-dimensional space mapping |
| CN112085869A (en) * | 2020-09-18 | 2020-12-15 | 陕西千山航空电子有限责任公司 | Civil aircraft flight safety analysis method based on flight parameter data |
| CN112965849B (en) * | 2021-03-05 | 2022-06-10 | 中国科学院微小卫星创新研究院 | Satellite fault diagnosis inference machine system and method |
| CN113570059A (en) * | 2021-07-21 | 2021-10-29 | 北京航天测控技术有限公司 | Spacecraft decision reasoning method, device and system |
| CN117742304B (en) * | 2024-02-09 | 2024-05-07 | 珠海市南特金属科技股份有限公司 | Fault diagnosis method and system for crankshaft double-top vehicle control system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101083019A (en) * | 2006-12-31 | 2007-12-05 | 中国人民解放军63791部队 | Rapid evaluating system based on roomage state sensing |
| CN101590918A (en) * | 2009-06-19 | 2009-12-02 | 上海微小卫星工程中心 | Method for automatic fault diagnosis of satellite and diagnostic system thereof |
| CN102495875A (en) * | 2011-12-02 | 2012-06-13 | 上海海洋大学 | Marine disaster early warning expert system based on data mining |
-
2015
- 2015-09-22 CN CN201510608592.XA patent/CN105159286B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101083019A (en) * | 2006-12-31 | 2007-12-05 | 中国人民解放军63791部队 | Rapid evaluating system based on roomage state sensing |
| CN101590918A (en) * | 2009-06-19 | 2009-12-02 | 上海微小卫星工程中心 | Method for automatic fault diagnosis of satellite and diagnostic system thereof |
| CN102495875A (en) * | 2011-12-02 | 2012-06-13 | 上海海洋大学 | Marine disaster early warning expert system based on data mining |
Non-Patent Citations (2)
| Title |
|---|
| 构建多航天器在轨管理支持平台;王环等;《航天器工程》;20070531;第16卷(第3期);第114-119页 * |
| 航天发射一体化建设与决策支持技术研究;王家伍;《装备指挥技术学院学报》;20060228;第17卷(第1期);第47页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105159286A (en) | 2015-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105159286B (en) | A kind of in-orbit abnormal alarm of spacecraft and fault diagnosis system | |
| Jouin et al. | Prognostics of PEM fuel cell in a particle filtering framework | |
| CN112749509B (en) | Intelligent substation fault diagnosis method based on LSTM neural network | |
| CN103926490A (en) | Power transformer comprehensive diagnosis method with self-learning function | |
| KR20110072746A (en) | Automated periodic surveillance testing method and apparatus in digital reactor protection system | |
| CN117394529A (en) | SCADA-based auxiliary decision method and system for main distribution network loop-closing reverse power supply control conditions | |
| CN109523030A (en) | A kind of telemetry parameter exception monitoring system based on machine learning | |
| CN112836843B (en) | Base station out-of-service alarm prediction method and device | |
| CN117523793A (en) | Power plant equipment fault early warning method and computer equipment | |
| CN108446202A (en) | A kind of judgment method of the safe condition of calculator room equipment | |
| CN105913226B (en) | Nuclear power plant's operation support system based on intelligent voice prompt | |
| CN114743703A (en) | Reliability analysis method, device, equipment and storage medium for nuclear power station unit | |
| CN115356990A (en) | Material yard equipment fault prediction method and system based on deep learning and storage medium | |
| Zavisca et al. | A bayesian network approach to accident management and estimation of source terms for emergency planning | |
| CN112966785A (en) | Intelligent constellation state identification method and system | |
| CN114091750A (en) | Power grid load abnormity prediction method, system and storage medium | |
| CN111210361A (en) | Power communication network routing planning method based on reliability prediction and particle swarm optimization | |
| Yamaguchi et al. | Data based construction of Bayesian network for fault diagnosis of event-driven systems | |
| US20230105839A1 (en) | Determining an action to allow resumption wind turbine operation after a stoppage | |
| CN116351545A (en) | LSTM-based coal mill fault early warning method | |
| Yu et al. | An online fault diagnosis method for nuclear power plant based on combined artificial neural network | |
| CN117038048B (en) | Remote fault processing method and system for medical instrument | |
| CN112801815B (en) | Power communication network fault early warning method based on federal learning | |
| CN117578742B (en) | Intelligent power dispatching system safety monitoring method and system | |
| US20230315940A1 (en) | Method and system for monitoring and/or operating a power system asset |
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 |