CN106325264B - A kind of the isolabilily evaluation method of UAV Flight Control System - Google Patents
A kind of the isolabilily evaluation method of UAV Flight Control System Download PDFInfo
- Publication number
- CN106325264B CN106325264B CN201610972338.2A CN201610972338A CN106325264B CN 106325264 B CN106325264 B CN 106325264B CN 201610972338 A CN201610972338 A CN 201610972338A CN 106325264 B CN106325264 B CN 106325264B
- Authority
- CN
- China
- Prior art keywords
- failure
- control system
- flight control
- uav flight
- vector
- 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
- G05B23/0259—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
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Geophysics And Detection Of Objects (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The present invention relates to a kind of the isolabilily evaluation methods of UAV Flight Control System.In order to realize the optimization design of UAV Flight Control System, in the complexity of unmanned aerial vehicle design stage needs assessment system failure separation.The present invention establishes UAV Flight Control System model according to UAV Flight Control System principle and fault type；Fault diagnosis residual generator is built using Parity space approach；Fault reconstruction condition is established using the COS distance of residual vector caused by different faults；Using the complexity of the similarity quantitative assessment fault reconstruction of residual vector, Distance conformability degree in pattern-recognition and direction similarity are combined, it is proposed that improve the isolabilily quantitative assessing index.The evaluation method can difference of the residual vector in distance and direction caused by overall merit different faults, provide a kind of reference frame for UAV system design and the design of fault reconstruction algorithm.
Description
Technical field
The present invention relates to a kind of the isolabilily evaluation methods of UAV Flight Control System, for evaluating unmanned plane
Actuator failures and sensor fault realize the complexity of fault reconstruction in flight control system, belong to UAV system failure
Separation technology field.
Background technology
Unmanned plane is used widely in fields such as modern times scouting, engineering mapping and scientific experiments, is had more considerable
Military and civilian foreground.UAV Flight Control System is that unmanned plane realizes autonomous flight or the control system of semi-autonomous flight,
Play the role of to stable and control UAV Attitude, management unmanned plane completion task vital.
Unmanned plane is easily influenced by complex environment factor in flight course, and event inevitably occurs for flight control system
Barrier, easily leads to unmanned plane and is difficult to complete task, or even cause unmanned plane crash accident.At present to UAV Flight Control System event
The research of barrier detection has obtained a large amount of achievements, but the design and failure of UAV Flight Control System fault reconstruction algorithm can divide
Wait further to study from property analysis.On the one hand, if not considering that failure can divide in the design phase of UAV Flight Control System
From property, the UAV Flight Control System for easily leading to design is difficult to realize fault reconstruction, and when system jam can not repair in time
Multiple failure；On the other hand, actuator failures and sensor fault lack unified failure and can divide in UAV Flight Control System
From property quantitative target, it is difficult to the complexity of quantitative assessment different faults separation, therefore unmanned plane the isolabilily is evaluated
Research it is most important.
Currently, for the system failure detach complexity evaluation problem, most of achievements in research according to control input with
Measure the redundancy relationship qualitative evaluation the isolabilily of output.The write paper such as Wang Zhenxi " based on system redundancy relationship can
Diagnostic methods research [J] Aerospace Controls, 2013,31 (6):Can 10-16,26 " cause output variable to become using failure variable
Change construction incidence matrix, analyzes the isolabilily using incidence matrix, but this method is only capable of in qualitative judgement system different events
Can barrier detach, and be unable to the complexity of quantitative assessment fault reconstruction, and fail to consider influence of the Unknown worm to fault reconstruction.
The write patent such as Wang great Yi " spacecraft control diagnosability determination method [P] China under a kind of influence of noise,
201410827895.6 " using Parity space approach generation fault diagnosis residual error, the evaluation problem of the isolabilily is converted
For residual vector Distance conformability degree discrimination or direction similarity discrimination, but this method be only capable of quantitative assessment residual error away from
From or direction on difference, fail difference of the overall merit residual vector in distance and direction.
In short, existing the isolabilily evaluation method be only capable of quantitative assessment different faults caused by residual vector in distance
On difference or the difference on direction, the isolabilily evaluation method proposed there has been no document or patent being capable of overall merit
Qualitative method and quantitative approach are simultaneously combined, the isolabilily evaluation method by difference of the residual vector in distance and direction
It needs to be studied.
Invention content
The technology of the present invention solves the problems, such as：The complexity of UAV Flight Control System fault reconstruction is evaluated, is utilized
The COS distance of fault diagnosis residual vector caused by different faults establishes fault reconstruction condition in equivalent space；Using residual error to
The complexity of the similarity quantitative assessment fault reconstruction of amount mutually ties Distance conformability degree in pattern-recognition with direction similarity
It closes, proposes a kind of improvement the isolabilily quantitative assessing index, quantitatively commented using fault reconstruction condition and the isolabilily
Valence metrics evaluation the isolabilily.
Technical solution of the invention is：A kind of the isolabilily evaluation method of UAV Flight Control System,
Include the following steps：
Step 1：According to UAV Flight Control System principle, unmanned plane linear discrete steady flight control system mould is established
Type is as follows：
Wherein,Respectively state variable, control
Input variable, output variable, Unknown worm variable and failure variable according to UAV Flight Control System structure and fly
Row environment determines；nx、nu、ny、nd、nfThe dimension of respectively x (k), u (k), y (k), d (k), f (k), k indicate sampling instant,All n are indicated respectivelyx、nu、ny、nd、nfTie up real vector；D (k) include noise, atmospheric perturbation with
And model uncertainty, it is assumed that Unknown worm variable is linear orthogonal zero-mean gaussian in UAV Flight Control System
Random vector Indicate mean value be 0, covariance matrix ΛdGaussian random distribution, 0 indicate zero moment
Battle array or null vector, ΛdFor the covariance matrix of d；Failure variable f is l in UAV Flight Control System2The determination of norm-bounded
Property additivity failure, andfi(k),1≤i≤nfFor i-th of component of failure variable f (k),
fi(k) ≠ 0 certain actuator or sensor failure are indicated；A, B, C, D are respectively the system square of UAV Flight Control System
Battle array, input matrix, output matrix and transmission matrix are determined according to the structure and parameter of UAV Flight Control System；Bf、DfFor
The known matrix or vector determined according to system failure f types, Bd、DdFor the known square determined according to system Unknown worm d types
Battle array or vector；
Step 2：According to UAV Flight Control System structure, parameter and computing capability, equivalent space exponent number s ∈ N are determined+, N+All positive integers are indicated, by the unmanned plane linear discrete steady flight control system control input variable u (k) described in step 1
With the redundancy relationship of output variable y (k), the equivalent equation of UAV Flight Control System is built in equivalent space；
Step 3：Change of equal value is carried out to equivalent equation in the equivalent space of the UAV Flight Control System described in step 2
It changes, is produced according to Unknown worm d and UAV Flight Control System parameter designing UAV Flight Control System fault diagnosis residual error
Raw device；
Step 4：Unknown worm is calculated according to the UAV Flight Control System fault diagnosis residual generator described in step 3
Residual vector r caused by ddWith different faults fiCaused nominal residue vector rfi,1≤i≤nf, according to rd、rfiCalculate different events
Hinder fiCaused residual vector ri,1≤i≤nf；
Step 5：According to nominal residue vector r caused by the different faults described in step 4fi, establish failure fiFailure inspection
Survey condition judges that can each failure meet fault detect condition, obtains detectable failure in UAV Flight Control System successivelyIndicate the detectable failure number of UAV Flight Control System；
Step 6：According in the UAV Flight Control System fault diagnosis residual generator and equivalent space described in step 3
Unknown worm variable dsProbability distribution, calculate residual vector r caused by Unknown worm ddProbability distribution, according to rdProbability
Distribution calculates the UAV Flight Control System detectable failure described in step 5Caused residual vectorProbability
Distribution；
Step 7：If failure is undetectable in UAV Flight Control SystemOr detectable failure numberIt is then
Failure pair to be separated is free of in system；If detectable failure numberAny two in UAV Flight Control System can detect
FailureConstruct failure pair to be separated；
Step 8：Since first group of failure to be separated to, according to the failure centering failure to be separated described in step 7
Caused residual vectorWith nominal residue vectorDirection difference, establish successively
Failure pair to be separatedFault reconstruction condition, judge failure pair to be separatedFault reconstruction condition can be met；
Step 9：If failure pair to be separatedMeet the fault reconstruction condition described in step 8, calculates failure pair to be separatedImprovement the isolabilily quantitative assessing indexJudge whether that all failure to be separated is to analysis
It finishes, if so, the isolabilily quantitative assessment terminates, otherwise returns to step 8 and continue to calculate, until all failures to be separated
The isolabilily evaluation is finished.
Wherein, the specific solution of equivalent equation of the UAV Flight Control System described in step 2 in equivalent space
Method is：First, according to equivalent space exponent number s, the equivalent equation of UAV Flight Control System is built such as in equivalent space
Under：
ys(k)-Husus(k)=Hosx(k-s)+Hdsds(k)+Hfsfs(k)
Wherein,
ys(k)、us(k)、ds(k)、fs(k) divide
It Biao Shi not output variable, control input variable, Unknown worm variable and the event of y (k), u (k), d (k), f (k) in equivalent space
Hinder variable；Indicate own (n respectivelyy·(s+1))
×(nu·(s+1))、(ny·(s+1))×(nd·(s+1))、(ny·(s+1))×(nf·(s+1))、(ny·(s+1))×nx
Real number matrix is tieed up, All n are indicated respectivelyu·(s+1)、nd·(s+1)、nf·(s+1)、
ny(s+1) real vector is tieed up.
Wherein, the specific design method of the UAV Flight Control System fault diagnosis residual generator described in step 3 is：
Equivalence transformation is carried out to the equivalent equation of UAV Flight Control System first, in equivalent equation equation both ends premultiplication
Matrix HosLeft null matrix Ns, i.e. NsHos=0, obtain the equivalent equation without state variable x (k-s)：
Ns(ys(k)-Husus(k))=NsHdsds(k)+NsHfsfs(k)
Then, N is calculatedsHdsdsCovariance matrixΛdsIndicate d in equivalent spacesAssociation
Variance matrix, and calculate P=(Λnds)-1/2, in above-mentioned equivalent equation both ends premultiplication matrix P to meetI indicates unit matrix, obtains equivalent equation：
PNs(ys(k)-Husus(k))=PNsHdsds(k)+PNsHfsfs(k)
Finally, design UAV Flight Control System fault diagnosis residual generator is：
R (k)=NosHdsds(k)+NosHfsfs(k)
Wherein, Nos=PNs, UAV Flight Control System fault diagnosis residual error r (k)=Nos(ys(k)-Husus(k))。
Wherein, the specific method for solving of residual vector caused by the different faults described in step 4 is：
According to UAV Flight Control System fault diagnosis residual generator r (k), calculate first failure do not occur f (k)=
Residual vector r caused by Unknown worm d (k) when 0d(k)=NosHdsds(k) and different faults fi(k) caused nominal when occurring
Residual vector rfi(k)=NosFifsi(k),1≤i≤nf, then calculate different faults fi(k) residual vector caused by, obtains：
ri(k)=NosFifsi(k)+NosHdsds(k),1≤i≤nf
Wherein,
Bf,i、Df,i,1≤i≤nfB is indicated respectivelyfI-th row, DfI-th row.
Wherein, it the fault detect condition described in step 5 and obtains the specific judgment method of detectable failure and is：
As failure fiWhen generation, i.e. fi(k) ≠ 0, fi(k) nominal residue vector r caused byfi(k)=NosFifsi(k)=0
Indicate failure fiGeneration cannot cause nominal residue vector rfiVariation, therefore failure fiResidual error r cannot be passed throughiIt is detected, fiNo
Meet fault detect condition；Work as fi(k) ≠ 0 when, if rfi(k) ≠ 0 failure fiGeneration can pass through residual error riIt is detected, fiIt is full
Sufficient fault detect condition；
It is right successivelyFailure judgement testing conditions obtain detectable failure in UAV Flight Control System Indicate the detectable failure number of UAV Flight Control System.
Wherein, the specific method for solving of the different residual vector probability distribution described in step 6 is：
First, residual vector r caused by Unknown worm d is calculateddProbability distribution；Due toAndResidual vector r caused by Unknown worm ddIt is equivalent to Unknown worm variable dsLinear transformation is carried out,
Obtain residual vector r caused by Unknown worm ddProbability distribution be
Then, detectable failure is calculatedCaused residual vectorProbability distribution；Due toCaused residual vector isNominal residue vector isWhereinObtain detectable failureCaused residual vector
Probability distribution it is as follows：
Wherein, It indicates respectively
Bf Row, Df Row.
Wherein, the specific method for solving of the failure pair to be separated described in step 7 is：
If institute is faulty undetectable in UAV Flight Control System, i.e.,Institute is faulty to can not achieve event
Barrier separation；If only there are one detectable failures in UAV Flight Control System, i.e.,It is former with other when then the failure occurs
Barrier separation problem is converted into the fault detection problem of the failure, and research the isolabilily is meaningless at this time, thereforeOrWhen UAV Flight Control System in be free of failure pair to be separated；
If detectable failure number in UAV Flight Control SystemBy any two detectable failureFailure pair to be separated is constituted, is obtainedGroup failure pair to be separated.
Wherein, the failure to be separated described in step 8 is to the specific judgment method of fault reconstruction condition：
First, detectable failure is establishedWhen generation and detectable failureRealize the cosine form of fault reconstruction condition；By
In detectable failureWhen generation and detectable failureRealize that the condition of fault reconstruction isCaused residual vectorWithDraw
The nominal residue vector risenBetween angle should be less thanWithCaused nominal residue vectorIts cosine form is：
Wherein,It indicatesEuropean norm, and so on；
Then, it will be based onCaused nominal residue vectorEuropean norm establishWhen generation and detectable failure
The condition abbreviation for realizing fault reconstruction, obtains：
Wherein,
Finally, fault reconstruction confidence alpha is determined according to the actual demand of UAV Flight Control System, due toIt establishes and separates failureWith confidence alpha and detectable failure when generationThe condition of separation is as follows：
Wherein：Failure pair to be separatedInFault reconstruction threshold value
It is distributed the confidence interval upper limit for the corresponding standard gaussian of confidence alpha；
Since first group of failure to be separated to, judgeCan group failure to be separated to meet fault reconstruction item
Part.
Wherein, the specific solution of the improvement the isolabilily quantitative assessing index of the failure pair to be separated described in step 9
Method is：
First, failure pair to be separated is calculatedCaused residual vectorK-L divergences improve indexThe index quantification evaluates residual errorDistance conformability degree, i.e.,：
Wherein,It indicatesrdK-L divergences (Kullback-Leibler divergence), due toIt is computed
It indicates respectivelyIn maximum value, minimum value；Due to Intentionally
Justice；
Then, failure pair to be separated is calculatedCaused nominal residue vectorSine valueThe index passes throughDirection difference quantitative assessment residual errorDirection similarity, i.e.,：
Further, it is contemplated that influence of the Unknown worm to fault reconstruction, introduces fault detectability quantitative assessing indexQuantitative assessment failure and Unknown worm cause the difference of the opposite variation degree of residual error, i.e.,：
Wherein：It indicatesIn minimum value, influence journey of the quantitative assessment failure to residual error
Degree；Failure determination threshold valueIt is determined by confidence alpha and residual variance,It is η's that expression confidence level, which is α degree of freedom,
Chi square distribution, η NosLine number, influence degree of the quantitative assessment Unknown worm to residual error；
Finally, failure pair to be separated is calculatedImprovement the isolabilily quantitative assessing index：
JudgeThe isolabilily of group failure pair to be separated evaluates whether that analysis finishes, if so, failure can
Separation property evaluation terminates, and otherwise returns to step 8 and continues to calculate, until all failure to be separated has evaluated the isolabilily
Finish.
The advantages of the present invention over the prior art are that：
(1) present invention establishes event according to the direction difference of fault diagnosis residual vector caused by different faults in equivalent space
Hinder separation condition, converts fault reconstruction condition to the European norm Yu fault reconstruction threshold value of different faults vector in equivalent space
Comparison problem, this method directly utilize the European norm failure judgement separation condition of fault vectors, more can intuitively evaluate
Can actuator failures and sensor fault realize fault reconstruction in UAV Flight Control System.
(2) present invention is separable using the similarity quantitative assessment failure of fault diagnosis residual vector caused by different faults
Distance conformability degree and direction similarity are combined by property, the improvement the isolabilily quantitative assessing index of proposition, can be more
Accurately difference of the residual vector in distance and direction caused by quantitative assessment different faults, improves the isolabilily and comments
The accuracy of valence, the isolabilily evaluation result are that UAV flight control system design and the design of fault reconstruction algorithm carry
For a kind of reference frame.
(3) the isolabilily qualitative evaluating method and quantitative evaluation method are combined by the present invention, compared with than existing failure
Separability evaluation method can more fully, accurately evaluate the complexity of UAV Flight Control System fault reconstruction.
Description of the drawings
Fig. 1 is UAV Flight Control System block diagram；
Fig. 2 is the isolabilily evaluation method flow chart of UAV Flight Control System proposed by the present invention；
Fig. 3 is UAV Flight Control System fault detect figure；
Fig. 4 is UAV Flight Control System fault reconstruction figure.
Specific implementation mode
UAV Flight Control System block diagram of the present invention is as shown in Figure 1, mainly by controller, actuator, unmanned plane body
It is formed with sensor, control instruction, in actuator, causes unmanned plane body position, speed by controller action by controlling input
The variation of the state variables such as degree and posture measure and will measure variable feedback to controller by sensor, and controller is again
According to control instruction and measure variable generation control input.The isolabilily evaluation method stream of UAV Flight Control System
Journey figure is as shown in Fig. 2, the specific method is as follows：
Step 1：According to UAV Flight Control System principle, unmanned plane linear discrete steady flight control system mould is established
Type is as follows：
Wherein,Respectively state variable, control
Input variable, output variable, Unknown worm variable and failure variable, according to UAV Flight Control System structure and flight environment of vehicle
It determines；nx、nu、ny、nd、nfThe dimension of respectively x (k), u (k), y (k), d (k), f (k), k indicate sampling instant,All n are indicated respectivelyx、nu、ny、nd、nfTie up real vector；D (k) include noise, atmospheric perturbation with
And model uncertainty, it is assumed that Unknown worm variable is linear orthogonal zero-mean gaussian in UAV Flight Control System
Random vector Indicate mean value be 0, covariance matrix ΛdGaussian random distribution, 0 indicate zero moment
Battle array or null vector, ΛdFor the covariance matrix of d；Failure variable f is l in UAV Flight Control System2The determination of norm-bounded
Property additivity failure, andfi(k),1≤i≤nfIt is i-th point of failure variable f (k)
Amount, fi(k) ≠ 0 certain actuator or sensor failure are indicated；A, B, C, D are respectively the system of UAV Flight Control System
Matrix, input matrix, output matrix and transmission matrix are determined according to the structure and parameter of UAV Flight Control System；Bf、Df
For the known matrix or vector determined according to system failure f types, Bd、DdKnown to being determined according to system Unknown worm d types
Matrix or vector；
Step 2：According to UAV Flight Control System structure, parameter and computing capability, equivalent space exponent number s ∈ N are determined+, N+All positive integers are indicated, by the unmanned plane linear discrete steady flight control system control input variable u (k) described in step 1
With the redundancy relationship of output variable y (k), the equivalent equation that UAV Flight Control System is built in equivalent space is as follows：
ys(k)-Husus(k)=Hosx(k-s)+Hdsds(k)+Hfsfs(k)
Wherein,
ys(k)、us(k)、ds(k)、fs(k)
Indicate respectively y (k), output variable in equivalent space of u (k), d (k), f (k), control input variable, Unknown worm variable and
Failure variable；Indicate own (n respectivelyy·(s+
1))×(nu·(s+1))、(ny·(s+1))×(nd·(s+1))、(ny·(s+1))×(nf·(s+1))、(ny·(s+1))
×nxReal number matrix is tieed up, All n are indicated respectivelyu·(s+1)、nd·(s+1)、nf·(s
+1)、ny(s+1) real vector is tieed up.
Step 3：Flown according to equivalent equation design unmanned plane in the UAV Flight Control System equivalent space described in step 2
Row control system fault diagnosis residual generator.First in the equivalent space of the UAV Flight Control System described in step 2
In equivalent equation equation both ends premultiplication matrix HosLeft null matrix Ns, i.e. NsHos=0, obtain without state variable x (k-s) etc.
Valence equation：Ns(ys(k)-Husus(k))=NsHdsds(k)+NsHfsfs(k), N then, is calculatedsHdsdsCovariance matrixΛdsIndicate d in equivalent spacesCovariance matrix, and calculate P=(Λnds)-1/2, above-mentioned etc.
Valence equation both ends premultiplication matrix P is to meetI indicates unit matrix, obtains equivalent equation：PNs(ys
(k)-Husus(k))=PNsHdsds(k)+PNsHfsfs(k), finally design UAV Flight Control System fault diagnosis residual error generates
Device：
R (k)=NosHdsds(k)+NosHfsfs(k)
Wherein, Nos=PNs, UAV Flight Control System fault diagnosis residual error r (k)=Nos(ys(k)-Husus(k))。
Step 4：Different faults are calculated according to the UAV Flight Control System fault diagnosis residual generator described in step 3
Caused residual error.Failure is calculated first, and residual vector r caused by Unknown worm d (k) when f (k)=0 does not occurd(k)=NosHdsds
(k) and different faults fi(k) caused nominal residue vector r when occurringfi(k)=NosFifsi(k),1≤i≤nf, then calculate
Different faults fi(k) residual vector caused by is：
ri(k)=NosFifsi(k)+NosHdsds(k),1≤i≤nf
Wherein,
Bf,i、Df,i,1≤i≤nfB is indicated respectivelyfI-th row, DfI-th row.
Step 5：According to nominal residue vector r caused by the different faults described in step 4fiFailure judgement testing conditions.When
Failure fiWhen generation, i.e. fi(k) ≠ 0, fi(k) nominal residue vector r caused byfi(k)=NosFifsi(k)=0 failure f is indicatedi's
Generation cannot cause nominal residue vector rfiVariation, therefore failure fiResidual error r cannot be passed throughiIt is detected, fiIt is unsatisfactory for fault detect
Condition；Work as fi(k) ≠ 0 when, if rfi(k) ≠ 0 failure fiGeneration can pass through residual error riIt is detected, fiMeet fault detect item
Part.
It is right successivelyFailure judgement testing conditions obtain detectable failure in UAV Flight Control System Indicate the detectable failure number of UAV Flight Control System.
Step 6：According to the probability distribution of UAV Flight Control System fault diagnosis residual generator and Unknown worm, meter
Calculate the probability distribution of residual error caused by UAV Flight Control System detectable failure.First, it calculates residual caused by Unknown worm d
Difference vector rdProbability distribution；Due toAndResidual vector r caused by Unknown worm dd
It is equivalent to Unknown worm variable dsLinear transformation is carried out, residual vector r caused by Unknown worm d is obtaineddProbability distribution be
Then, detectable failure f is calculatediCaused residual vector riProbability distribution.Due to fiCaused residual vector isNominal residue vector isWhereinObtain detectable failureCaused residual vector
Probability distribution it is as follows：
Wherein, It indicates respectively
Bf Row, Df Row.
Step 7：Waiting for point for UAV Flight Control System is constructed according to the detectable failure of UAV Flight Control System
From failure pair.If institute is faulty undetectable in UAV Flight Control System, i.e.,Institute is faulty to can not achieve event
Barrier separation；If only there are one detectable failures in UAV Flight Control System, i.e.,It is former with other when then the failure occurs
Barrier separation problem is converted into the fault detection problem of the failure, and research the isolabilily is meaningless at this time, thereforeOrWhen UAV Flight Control System in be free of failure pair to be separated.
If detectable failure number in UAV Flight Control SystemBy any two detectable failureFailure pair to be separated is constituted, is obtainedGroup failure pair to be separated.
Step 8：The fault reconstruction condition for establishing failure pair to be separated successively, judges that UAV Flight Control System is to be separated
Can failure to meet fault reconstruction condition.First, detectable failure is establishedWhen generation and detectable failureRealize failure
The cosine form of separation condition；Due to detectable failureWhen generation and detectable failureRealize that the condition of fault reconstruction is
Caused residual vectorWithCaused nominal residue vectorBetween angle should be less thanWithCaused nominal residue vectorIts cosine form is：
Wherein,It indicatesEuropean norm, and so on.
Then, it will be based onCaused nominal residue vectorEuropean norm establishWhen generation and detectable failure
The condition abbreviation for realizing fault reconstruction, obtains：
Wherein,
Finally, fault reconstruction confidence alpha is determined according to the actual demand of UAV Flight Control System, due toIt establishes and separates failureWith confidence alpha and detectable failure when generationThe condition of separation is as follows：
Wherein：Failure pair to be separatedInFault reconstruction threshold value
It is distributed the confidence interval upper limit for the corresponding standard gaussian of confidence alpha.
Since first group of failure to be separated to, judgeCan group failure to be separated to meet fault reconstruction item
Part.
Step 9：If failure pair to be separatedMeet the fault reconstruction condition described in step 8, calculates failure pair to be separatedImprovement the isolabilily quantitative assessing index.First, failure pair to be separated is calculatedCaused residual vectorK-L divergences improve indexThe index quantification evaluates residual errorDistance conformability degree, i.e.,：
Wherein,It indicatesrdK-L divergences (Kullback-Leibler divergence), due toIt is computed
It indicates respectivelyIn maximum value, minimum value；Due to Intentionally
Justice.
Then, failure pair to be separated is calculatedCaused nominal residue vectorSine valueThe index passes throughDirection difference quantitative assessment residual errorDirection similarity, i.e.,：
Further, it is contemplated that influence of the Unknown worm to fault reconstruction, introduces fault detectability quantitative assessing indexQuantitative assessment failure and Unknown worm cause the difference of the opposite variation degree of residual error, i.e.,：
Wherein：It indicatesIn minimum value, influence journey of the quantitative assessment failure to residual error
Degree；Failure determination threshold valueIt is determined by confidence alpha and residual variance,It is η's that expression confidence level, which is α degree of freedom,
Chi square distribution, η NosLine number, influence degree of the quantitative assessment Unknown worm to residual error.
Finally, failure pair to be separated is calculatedImprovement the isolabilily quantitative assessing index：
JudgeThe isolabilily of group failure pair to be separated evaluates whether that analysis finishes, if so, failure can
Separation property evaluation terminates, and otherwise returns to step 8 and continues to calculate, until all failure to be separated has evaluated the isolabilily
Finish.
The present invention is with a kind of fixed-wing unmanned plane in H0=7000m, V0The constant speed of=0.4Ma is without sideslip straight and level flight state
Under longitudinal model for evaluate the isolabilily.
Step 1：Take system state variables x (t)=[V α q θ H]T, control input quantity u (t)=[δe δp]T, measure and become
Measure y (t)=[V q θ H]T, f (t) indicates actuator failures or sensor fault, wherein actuator failures in the present embodiment
With elevator perseverance deviation fault f1For=0.5, sensor fault is with pitch gyro perseverance deviation fault f2=0.5, vertical
Gyro perseverance deviation fault f3For=0.5, fault vectors f=[f1 f2 f3]T, it is assumed that Unknown wormWherein d1、d2And d3Respectively process noise, atmospheric interference and measurement noise, wherein
Atmospheric interference d2Middle consideration horizontal wind speed and vertical velocity, ignore lateral wind speed, establish unmanned plane LINEAR CONTINUOUS steady flight control
System model processed is as follows：
Wherein：
Under Matlab simulated environment, sampling time T=0.01s is enabled, using Matlab discretization tools, establishes unmanned plane
Linear discrete steady flight control system model is as follows：
Wherein：
Step 2：According to UAV Flight Control System state dimension nx=5, set equivalent space exponent number s=5, confidence level
It is as follows to establish equivalent equation in equivalent space for α=0.05：
ys(k)-Husus(k)=Hosx(k-s)+Hdsds(k)+Hfsfs(k)
Wherein, us(k)=[uT(k-5) uT(k-4) … uT(k)]T, ys(k)、ds(k)、fs(k) and so on；
Hus、HdsAnd so on.
Step 3：H is calculated firstosLeft null matrix
In equivalent equation equation both ends premultiplication matrix Hos, obtain：
Ns(ys(k)-Husus(k))=NsHdsds(k)+NsHfsfs(k)
Then, N is calculatedosHdsds(k) covariance matrix
And it calculatesObtain equivalent equation：
PNs(ys(k)-Husus(k))=PNsHdsds(k)+PNsHfsfs(k)
Finally, design UAV Flight Control System fault diagnosis residual generator is：
R (k)=NosHdsds(k)+NosHfsfs(k)
Wherein,
Step 4：According to the UAV Flight Control System fault diagnosis residual generator r (k) described in step 3, r is obtainedd
(k)=NosHdsds(k), different faults fiCaused nominal residue vector r when generationfi(k)=NosFifsi(k) (i=1,2,3)
And different faults fiCaused residual vector r when generationi(k)=NosFifsi(k)+NosHdsds(k) (i=1,2,3), wherein fsi
(k)=[0.5 0.5 ... 0.5]T∈R6(i=1,2,3),F2、
F3And so on.
Step 5：According to nominal residue vector r caused by the different faults described in step 4fi, due to fi(k) ≠ 0 (i=1,
2,3) when, rf1(k)=[- 0.2752 1.0864-1.6123-0.0190 ... 3.1002]T∈R19≠ 0, obtain f1Meet failure
Separation condition can similarly obtain f2、f3Meet fault reconstruction condition, f1、f2、f3It is detectable failure.
Step 6：It calculatesCause residual errorProbability distribution be
Step 7：Due to detectable failure number in UAV Flight Control SystemBy the detectable event of any two
Barrier constitutes failure pair to be separated, and 3 groups of failures pair to be separated are obtained.
Step 8：According toCaused nominal residue vectorEuropean norm, establish f1When generation and f2The condition of separation
It is as follows：
Wherein, fs1,1=0.4082 × [1 1111 1]T, rf1,1=[- 0.2247 0.8870-1.3165 ...
2.5313]T∈R19, n1=[- 0.0566 0.2234-0.3316 ... 0.6375]T∈R19, n2=[- 0.3592 0.3097
0.0402 … 0.0118]T∈R19, n12=[- 0.4158 0.5331-0.2913 ... 0.6493]T∈R19。
According to confidence level 0.05, due toObtain ρ12=1.96, it establishes and separates failure f1When generation with
Confidence level 0.05 and detectable failure f2The condition of separation is | | fs1||2＞ f12,1c=0.7305.
Due to | | fs1||2=1.2247 ＞ 0.7305, f1It can be with confidence level 0.05 and detectable failure f when generation2Separation,
Remaining fault reconstruction, it is as shown in table 1 to obtain fault reconstruction threshold value, wherein failure f1It cannot be with confidence level when generation
0.05 and detectable failure f3Separation, remaining situation can realize fault reconstruction.
1 unmanned plane Longitudinal Control System fault reconstruction threshold value of table
Step 9：First, failure to be separated is calculated to f1、f2Caused residual vector r1、r2K-L divergences improve index：Then failure to be separated is calculated to f1、f2Caused nominal residue vector
rf1、rf2Sine value D2(r1,r2)=sin2(rf1,rf2)=0.9962；0.05 time failure determination threshold value J of confidence levelth=
30.14, calculate fault detectability quantitative assessing index：Obtain failure to be separated
To f1、f2Improvement the isolabilily quantitative assessing index D1,2=246.8, the improvement failure of other failures pair to be separated can divide
From property quantitative assessing index, obtain improving the isolabilily quantitative assessing indexAs shown in table 2.
2 unmanned plane Longitudinal Control System of table improves the isolabilily quantitative assessing index
As shown in Table 2, failure to be separated is to f1、f3Improvement the isolabilily quantitative assessing index D1,3Far below other
The improvement the isolabilily quantitative assessing index of failure pair to be separated, is consistent with 1 result of table.
In order to verify the analysis result of the present invention, fault detect is carried out to above-mentioned UAV Flight Control System first, is imitated
True 5s data, f is sequentially added in 2s1、f2、f3, fault detect residual error evaluation function is respectively Jk1, Jk2, Jk3；Confidence level
0.05 time failure determination threshold value is Indicate confidence level be 0.05, the card side that degree of freedom is 19
Distribution, it is as shown in Figure 3 to obtain UAV Flight Control System fault detect figure.Due to failure f in Fig. 31、f2、f3Failure after generation
Residual error evaluation function Jk1, Jk2, Jk3 is detected obviously higher than failure determination threshold value Jth, illustrates failure f1、f2、f3It is detectable.
Utilization orientation residual error of the present invention carries out the correctness that fault reconstruction verifies this evaluation method, calculates successively to be separated
Residual error riWith nominal residue rf1、rf2、rf3COS distance, due to the maximum nominal residue of COS distance and riDirection is most close,
Phylogenetic fault type is determined by the fault type of the nominal residue.Remember failure fiCaused riWith fjNominal residue to
Measure rfjCOS distance be Jij, UAV Flight Control System fault reconstruction figure is as shown in figure 4, wherein failure f1J when generation11、
J12、J13As shown in fig. 4 a；Failure f2J when generation21、J22、J23As shown in Figure 4 b；Failure f3J when generation31、J32、J33Such as Fig. 4 c institutes
Show, second figure is the partial enlarged view of first figure in Fig. 4 c.
By Fig. 4 a it is found that f1When generation, failure f1Caused residual vector r1With failure f1Nominal residue vector rf1It is remaining
Chordal distance J11Significantly greater than r1With failure f2Nominal residue vector rf2COS distance J12, illustrate failure f1When generation residual error to
Measure r1With rf1Angle be significantly less than r1With rf2Angle, f1It is easy to when generation and f2Realize fault reconstruction；But r1With rf1Cosine
Distance J11Same r1With rf3COS distance J13There is no significant difference, illustrates failure f1R when generation1With rf1The same r of angle1With rf3's
Angle does not have significant difference, f1It is difficult to when generation and f3It realizes fault reconstruction, is consistent with evaluation method of the present invention.By Fig. 4 b it is found that
f2When generation, r2With rf2COS distance J22Significantly greater than r2With rf1、rf3COS distance J21、J23, illustrate failure f2When generation
r2With rf2Angle be significantly less than r2With rf1、rf3Angle, f2It is easy to when generation and f1、f3Fault reconstruction is realized, with institute of the present invention
Evaluation method is stated to be consistent.By Fig. 4 c it is found that f3When generation, r3With rf3COS distance J33Slightly larger than r3With rf1COS distance
J31, illustrate r3With rf3Angle be slightly less than r3With rf1Angle, f3Energy and f when generation1Realize fault reconstruction；r3With rf3Cosine
Distance J33Significantly greater than r3With rf2COS distance J32, illustrate r3With rf3Angle be significantly less than r3With rf2Angle, f3Occur
When be easy to and f2It realizes fault reconstruction, is consistent with evaluation method of the present invention, demonstrates effectiveness of the invention.
Claims (9)
1. a kind of the isolabilily evaluation method of UAV Flight Control System, it is characterised in that：Include the following steps：
Step 1：According to UAV Flight Control System principle, unmanned plane linear discrete steady flight control system model is established such as
Under：
Wherein,Respectively state variable, control input
Variable, output variable, Unknown worm variable and failure variable, according to UAV Flight Control System structure and flying ring
Border determines；nx、nu、ny、nd、nfThe dimension of respectively x (k), u (k), y (k), d (k), f (k), k indicate sampling instant,All n are indicated respectivelyx、nu、ny、nd、nfTie up real vector；D (k) include noise, atmospheric perturbation with
And model uncertainty, it is assumed that Unknown worm variable is linear orthogonal zero-mean gaussian in UAV Flight Control System
Random vector Indicate mean value be 0, covariance matrix ΛdGaussian random distribution, 0 indicate zero moment
Battle array or null vector, ΛdFor the covariance matrix of d；Failure variable f is l in UAV Flight Control System2The determination of norm-bounded
Property additivity failure, andfi(k),1≤i≤nfFor i-th of component of failure variable f (k),
fi(k) ≠ 0 certain actuator or sensor failure are indicated；A, B, C, D are respectively the system square of UAV Flight Control System
Battle array, input matrix, output matrix and transmission matrix are determined according to the structure and parameter of UAV Flight Control System；Bf、DfFor
The known matrix or vector determined according to system failure f types, Bd、DdFor the known square determined according to system Unknown worm d types
Battle array or vector；
Step 2：According to UAV Flight Control System structure, parameter and computing capability, equivalent space exponent number s ∈ N are determined+, N+Table
Show all positive integers, by the unmanned plane linear discrete steady flight control system control input variable u (k) and defeated described in step 1
The redundancy relationship for going out variable y (k) builds the equivalent equation of UAV Flight Control System in equivalent space；
Step 3：Equivalence transformation, root are carried out to equivalent equation in the equivalent space of the UAV Flight Control System described in step 2
According to Unknown worm d and UAV Flight Control System parameter designing UAV Flight Control System fault diagnosis residual generator；
Step 4：Unknown worm d is calculated according to the UAV Flight Control System fault diagnosis residual generator described in step 3 to draw
The residual vector r risendWith different faults fiCaused nominal residue vector rfi,1≤i≤nf, according to rd、rfiCalculate different faults
fiCaused residual vector ri,1≤i≤nf；
Step 5：According to nominal residue vector r caused by the different faults described in step 4fi, establish failure fiFault detect item
Part judges that can each failure meet fault detect condition, obtains detectable failure in UAV Flight Control System successivelyIndicate the detectable failure number of UAV Flight Control System；
Step 6：According to unknown in the UAV Flight Control System fault diagnosis residual generator and equivalent space described in step 3
Input variable dsProbability distribution, calculate residual vector r caused by Unknown worm ddProbability distribution, according to rdProbability distribution
Calculate the UAV Flight Control System detectable failure described in step 5Caused residual vectorProbability point
Cloth；
Step 7：If failure is undetectable in UAV Flight Control SystemOr detectable failure numberThen in system
Without failure pair to be separated；If detectable failure numberBy any two detectable failure in UAV Flight Control SystemConstruct failure pair to be separated；
Step 8：Since first group of failure to be separated to, according to the failure centering failure to be separated described in step 7Cause
Residual vectorWith nominal residue vectorDirection difference, establish wait for point successively
From failure pairFault reconstruction condition, judge failure pair to be separatedFault reconstruction condition can be met；
Step 9：If failure pair to be separatedMeet the fault reconstruction condition described in step 8, calculates failure pair to be separated
Improvement the isolabilily quantitative assessing indexJudge whether that all failure to be separated is to having analyzed
Finish, if so, the isolabilily quantitative assessment terminates, otherwise return to step 8 and continue to calculate, until all failures pair to be separated
The isolabilily evaluation finishes.
2. the isolabilily evaluation method of UAV Flight Control System according to claim 1, it is characterised in that：
The specific method for solving of equivalent equation of the UAV Flight Control System in equivalent space described in step 2 is：
First, according to equivalent space exponent number s, the equivalent equation that UAV Flight Control System is built in equivalent space is as follows：
ys(k)-Husus(k)=Hosx(k-s)+Hdsds(k)+Hfsfs(k)
Wherein, ys(k)、us(k)、ds(k)、fs(k) y is indicated respectively
(k), output variable, control input variable, Unknown worm variable and the failure variable of u (k), d (k), f (k) in equivalent space；Indicate own (n respectivelyy·(s+1))×(nu·
(s+1))、(ny·(s+1))×(nd·(s+1))、(ny·(s+1))×(nf·(s+1))、(ny·(s+1))×nxTie up real number
Matrix, All n are indicated respectivelyu·(s+1)、nd·(s+1)、nf·(s+1)、ny·(s
+ 1) real vector is tieed up.
3. the isolabilily evaluation method of UAV Flight Control System according to claim 2, it is characterised in that：
The specific design method of UAV Flight Control System fault diagnosis residual generator described in step 3 is：
Equivalence transformation is carried out to the equivalent equation of UAV Flight Control System first, in equivalent equation equation both ends premultiplication matrix
HosLeft null matrix Ns, i.e. NsHos=0, obtain the equivalent equation without state variable x (k-s)：
Ns(ys(k)-Husus(k))=NsHdsds(k)+NsHfsfs(k)
Then, N is calculatedsHdsdsCovariance matrixΛdsIndicate d in equivalent spacesCovariance square
Battle array, and calculate P=(Λnds)-1/2, in above-mentioned equivalent equation both ends premultiplication matrix P to meet
It indicates unit matrix, obtains equivalent equation：
PNs(ys(k)-Husus(k))=PNsHdsds(k)+PNsHfsfs(k)
Finally, design UAV Flight Control System fault diagnosis residual generator is：
R (k)=NosHdsds(k)+NosHfsfs(k)
Wherein, Nos=PNs, UAV Flight Control System fault diagnosis residual error r (k)=Nos(ys(k)-Husus(k))。
4. the isolabilily evaluation method of UAV Flight Control System according to claim 3, it is characterised in that：
The specific method for solving of residual vector caused by different faults described in step 4 is：
According to UAV Flight Control System fault diagnosis residual generator r (k), when f (k)=0 does not occur for calculating failure first
Residual vector r caused by Unknown worm d (k)d(k)=NosHdsds(k) and different faults fi(k) caused nominal residue when occurring
Vectorial rfi(k)=NosFifsi(k),1≤i≤nf, then calculate different faults fi(k) residual vector caused by, obtains：
ri(k)=NosFifsi(k)+NosHdsds(k),1≤i≤nf
Wherein,
Bf,i、Df,i,1≤i≤nfB is indicated respectivelyfI-th row, DfI-th row.
5. the isolabilily evaluation method of UAV Flight Control System according to claim 4, it is characterised in that：
Fault detect condition described in the step 5 and specific judgment method for obtaining detectable failure is：
As failure fiWhen generation, i.e. fi(k) ≠ 0, fi(k) nominal residue vector r caused byfi(k)=NosFifsi(k)=0 event is indicated
Hinder fiGeneration cannot cause nominal residue vector rfiVariation, therefore failure fiResidual error r cannot be passed throughiIt is detected, fiIt is unsatisfactory for event
Hinder testing conditions；Work as fi(k) ≠ 0 when, if rfi(k) ≠ 0 failure fiGeneration can pass through residual error riIt is detected, fiMeet failure
Testing conditions；
It is right successivelyFailure judgement testing conditions obtain detectable failure in UAV Flight Control System Indicate the detectable failure number of UAV Flight Control System.
6. the isolabilily evaluation method of UAV Flight Control System according to claim 4, it is characterised in that：
The specific method for solving of different residual vector probability distribution described in step 6 is：
First, residual vector r caused by Unknown worm d is calculateddProbability distribution；Due toAndResidual vector r caused by Unknown worm ddIt is equivalent to Unknown worm variable dsLinear transformation is carried out,
Obtain residual vector r caused by Unknown worm ddProbability distribution be
Then, detectable failure is calculatedCaused residual vectorProbability distribution；Due toCaused residual vector isNominal residue vector isWhereinObtain detectable failureCaused residual vector
Probability distribution it is as follows：
Wherein, B is indicated respectivelyf Row, Df Row.
7. the isolabilily evaluation method of UAV Flight Control System according to claim 1, it is characterised in that：
The specific method for solving of failure pair to be separated described in step 7 is：
If institute is faulty undetectable in UAV Flight Control System, i.e.,Institute is faulty to can not achieve failure point
From；If only there are one detectable failures in UAV Flight Control System, i.e.,Divide with other failures when then the failure occurs
From the fault detection problem that problem is converted into the failure, research the isolabilily is meaningless at this time, thereforeOr
When UAV Flight Control System in be free of failure pair to be separated；
If detectable failure number in UAV Flight Control SystemBy any two detectable failure
Failure pair to be separated is constituted, is obtainedGroup failure pair to be separated.
8. the isolabilily evaluation method of UAV Flight Control System according to claim 6, it is characterised in that：
Failure to be separated described in step 8 is to the specific judgment method of fault reconstruction condition：
First, detectable failure is establishedWhen generation and detectable failureRealize the cosine form of fault reconstruction condition；Due to can
Detect failureWhen generation and detectable failureRealize that the condition of fault reconstruction isCaused residual vectorWithIt is caused
Nominal residue vectorBetween angle should be less thanWithCaused nominal residue vectorIts cosine form is：
Wherein,It indicatesEuropean norm, and so on；
Then, it will be based onCaused nominal residue vectorEuropean norm establishWhen generation and detectable failureIt realizes
The condition abbreviation of fault reconstruction, obtains：
Wherein,
Finally, fault reconstruction confidence alpha is determined according to the actual demand of UAV Flight Control System, due toIt establishes and separates failureWith confidence alpha and detectable failure when generationThe condition of separation is as follows：
Wherein：Failure pair to be separatedInFault reconstruction threshold valueFor
The corresponding standard gaussian of confidence alpha is distributed the confidence interval upper limit；
Since first group of failure to be separated to, judgeCan group failure to be separated to meet fault reconstruction condition.
9. the isolabilily evaluation method of UAV Flight Control System according to claim 8, it is characterised in that：
The specific method for solving of the improvement the isolabilily quantitative assessing index of failure pair to be separated described in step 9 is：
First, failure pair to be separated is calculatedCaused residual vectorK-L divergences improve indexThe index quantification evaluates residual errorDistance conformability degree, i.e.,：
Wherein,It indicatesK-L divergences (Kullback-Leibler divergence), due toIt is computed
It indicates respectivelyIn maximum value, minimum value；Due to Intentionally
Justice；
Then, failure pair to be separated is calculatedCaused nominal residue vectorSine valueThe index passes throughDirection difference quantitative assessment residual errorDirection similarity, i.e.,：
Further, it is contemplated that influence of the Unknown worm to fault reconstruction, introduces fault detectability quantitative assessing indexQuantitative assessment failure and Unknown worm cause the difference of the opposite variation degree of residual error, i.e.,：
Wherein：It indicatesIn minimum value, influence degree of the quantitative assessment failure to residual error；Therefore
Hinder detection threshold valueIt is determined by confidence alpha and residual variance,Indicate that confidence level is the card side point that α degree of freedom is η
Cloth, η NosLine number, influence degree of the quantitative assessment Unknown worm to residual error；
Finally, failure pair to be separated is calculatedImprovement the isolabilily quantitative assessing index：
JudgeThe isolabilily of group failure pair to be separated evaluates whether that analysis finishes, if so, failure is separable
Property evaluation terminate, otherwise returning to step 8 continues to calculate, until all failures to be separated finish the isolabilily evaluation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610972338.2A CN106325264B (en) | 2016-11-04 | 2016-11-04 | A kind of the isolabilily evaluation method of UAV Flight Control System |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610972338.2A CN106325264B (en) | 2016-11-04 | 2016-11-04 | A kind of the isolabilily evaluation method of UAV Flight Control System |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106325264A CN106325264A (en) | 2017-01-11 |
| CN106325264B true CN106325264B (en) | 2018-08-14 |
Family
ID=57816607
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610972338.2A Active CN106325264B (en) | 2016-11-04 | 2016-11-04 | A kind of the isolabilily evaluation method of UAV Flight Control System |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106325264B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106959700B (en) * | 2017-03-21 | 2019-08-27 | 北京航空航天大学 | A kind of unmanned aerial vehicle group collaboration patrol tracing path planing method based on upper limit confidence interval algorithm |
| CN107340764B (en) * | 2017-06-08 | 2018-11-06 | 腾讯科技（深圳）有限公司 | Abnormality eliminating method, device, storage medium and the unmanned plane terminal of unmanned plane |
| CN107356282A (en) * | 2017-06-23 | 2017-11-17 | 山东科技大学 | Bullet train robust interval Transducer-fault Detecting Method in the case of resolution limitations |
| CN107272648B (en) * | 2017-07-03 | 2019-05-03 | 山东科技大学 | A kind of separation method of industrial process small fault |
| CN107450320B (en) * | 2017-08-22 | 2020-09-11 | 广州大学 | Fuzzy self-adaptive compensation control method for actuator failure |
| CN107608335B (en) * | 2017-09-14 | 2020-02-14 | 山东科技大学 | Data driving method for fault detection and fault separation of unmanned aerial vehicle flight control system |
| CN108170127B (en) * | 2017-12-28 | 2019-07-26 | 山东科技大学 | A kind of fault detection method of UAV Flight Control System |
| CN108181917B (en) * | 2018-01-02 | 2021-07-13 | 佛山科学技术学院 | Spacecraft attitude control system fault diagnosability quantitative analysis method |
| CN111259553B (en) * | 2020-01-19 | 2020-10-23 | 北京空间飞行器总体设计部 | Spacecraft system fault detection obtaining method based on distance similarity |
| CN111308985B (en) * | 2020-02-18 | 2021-03-26 | 北京航空航天大学 | Performance degradation evaluation method for control assembly of airplane environmental control system based on NSCT and DM |
| CN112907781B (en) * | 2021-02-07 | 2022-08-19 | 中国人民解放军国防科技大学 | System fault diagnosis method and device, computer equipment and storage medium |
| CN114545907B (en) * | 2022-03-15 | 2023-12-19 | 中南大学 | Fault detection method of flight control system based on filter |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8683400B1 (en) * | 2012-11-21 | 2014-03-25 | Cadence Design Systems, Inc. | System and method for fault sensitivity analysis of mixed-signal integrated circuit designs |
| CN103697915A (en) * | 2013-12-24 | 2014-04-02 | 北京控制工程研究所 | Diagnostic evaluation method considering disturbing influence for satellite sensor fault |
| CN104571087A (en) * | 2014-12-26 | 2015-04-29 | 北京控制工程研究所 | Diagnostic determination method for spacecraft control system under influence of noise |
| CN105068529A (en) * | 2015-08-05 | 2015-11-18 | 清华大学 | Fault isolation method and system based on weighted residual component comparison |
| WO2015185070A1 (en) * | 2014-06-02 | 2015-12-10 | Volvo Truck Corporation | A method and system for monitoring the status of battery cells |
-
2016
- 2016-11-04 CN CN201610972338.2A patent/CN106325264B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8683400B1 (en) * | 2012-11-21 | 2014-03-25 | Cadence Design Systems, Inc. | System and method for fault sensitivity analysis of mixed-signal integrated circuit designs |
| CN103697915A (en) * | 2013-12-24 | 2014-04-02 | 北京控制工程研究所 | Diagnostic evaluation method considering disturbing influence for satellite sensor fault |
| WO2015185070A1 (en) * | 2014-06-02 | 2015-12-10 | Volvo Truck Corporation | A method and system for monitoring the status of battery cells |
| CN104571087A (en) * | 2014-12-26 | 2015-04-29 | 北京控制工程研究所 | Diagnostic determination method for spacecraft control system under influence of noise |
| CN105068529A (en) * | 2015-08-05 | 2015-11-18 | 清华大学 | Fault isolation method and system based on weighted residual component comparison |
Non-Patent Citations (4)
| Title |
|---|
| Generalization and analysis of sufficient conditions for PCA-based fault detectability and isolability;Baligh Mnassri 等;《Annual Reviews in Control》;20131231(第37期);第154-162页 * |
| Optimal Actuator Fault Detection for a TGSCM System Based on Disturbance Compensation;Maiying Zhong 等;《IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS》;20160630;第63卷(第5期);第3205-3215页 * |
| 动态系统实际故障可诊断性的量化评价研究;李文博 等;《自动化学报》;20150331;第41卷(第3期);第497-507页 * |
| 定量与定性相结合的动量轮故障可诊断性评价;刘文静 等;《中国空间科学技术》;20110831(第4期);第54-63页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106325264A (en) | 2017-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106325264B (en) | A kind of the isolabilily evaluation method of UAV Flight Control System | |
| CN106200629B (en) | A kind of detectable degree analysis method of the failure of UAV Flight Control System | |
| CN108196532B (en) | Fault detection and separation method for longitudinal flight control system of unmanned aerial vehicle based on nonlinear adaptive observer | |
| US20160114903A1 (en) | Flight envelope protection system for unmanned aerial vehicles | |
| CN103697915B (en) | A kind of satellite sensor failure diagnosticability evaluation method of considering interference effect | |
| CN107421534A (en) | A kind of redundance type SINS multiple faults partition method | |
| CN110554606B (en) | Self-adaptive fault-tolerant control method for hypersonic aircraft | |
| Klein et al. | Compatibility check of measured aircraft responses using kinematic equations and extended Kalman filter | |
| CN104571087B (en) | Spacecraft control diagnosability determination method under a kind of influence of noise | |
| Jain et al. | Model-free reconfiguration mechanism for fault tolerance | |
| CN102162733A (en) | Method for correcting autonomous underwater vehicle (AUV) dead reckoning navigation error in real time based on space vector modulation (SVM) | |
| CN108592946A (en) | A kind of online monitoring method of inertia device drift based under two sets of rotation inertial navigation redundant configurations | |
| D'Amato et al. | Fault tolerant low cost IMUS for UAVs | |
| CN103699121B (en) | Analytical redundancy relationship-based diagnosability determination method for satellite control system sensors | |
| Lerro et al. | Model-free scheme for angle-of-attack and angle-of-sideslip estimation | |
| Fravolini et al. | Experimental evaluation of two pitot free analytical redundancy techniques for the estimation of the airspeed of an UAV | |
| Liu et al. | An IMU fault diagnosis and information reconstruction method based on analytical redundancy for autonomous underwater vehicle | |
| Schettini et al. | Novel approach for angles calibration of air-data systems with inertial measurements | |
| CN107014386A (en) | The disturbing acceleration measuring method that a kind of attitude of flight vehicle is resolved | |
| CN104408326B (en) | A kind of appraisal procedure to survey of deep space independent navigation filtering algorithm | |
| CN104316052B (en) | A kind of nine gyro MEMS inertial measuring units based on orthogonal rotation | |
| Bittner et al. | Fault detection, isolation, and recovery techniques for large clusters of inertial measurement units | |
| Lombaerts et al. | Design and flight testing of nonlinear autoflight control laws incorporating direct lift control | |
| Saderla et al. | Parameter estimation of UAV from flight data using neural network | |
| Sevil | Airdata sensor based position estimation and fault diagnosis in aerial refueling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |