CN115048959A - RMSD-DS-based gun recoil prevention device fault diagnosis method - Google Patents
RMSD-DS-based gun recoil prevention device fault diagnosis method Download PDFInfo
- Publication number
- CN115048959A CN115048959A CN202210649020.6A CN202210649020A CN115048959A CN 115048959 A CN115048959 A CN 115048959A CN 202210649020 A CN202210649020 A CN 202210649020A CN 115048959 A CN115048959 A CN 115048959A
- Authority
- CN
- China
- Prior art keywords
- fault
- evidence
- rmsd
- recoil
- evidences
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 76
- 238000000034 method Methods 0.000 title claims abstract description 60
- 239000008186 active pharmaceutical agent Substances 0.000 title claims abstract description 29
- 230000002265 prevention Effects 0.000 title claims abstract description 24
- 230000004927 fusion Effects 0.000 claims abstract description 27
- 230000010354 integration Effects 0.000 claims abstract description 19
- 238000012935 Averaging Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 39
- 238000005299 abrasion Methods 0.000 claims description 32
- 239000011159 matrix material Substances 0.000 claims description 30
- 238000006073 displacement reaction Methods 0.000 claims description 20
- 238000005315 distribution function Methods 0.000 claims description 10
- 230000007547 defect Effects 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 208000037408 Device failure Diseases 0.000 claims 2
- 238000002405 diagnostic procedure Methods 0.000 claims 2
- 230000007257 malfunction Effects 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 5
- 238000013528 artificial neural network Methods 0.000 description 6
- 238000007500 overflow downdraw method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Locating Faults (AREA)
Abstract
The invention discloses a fault diagnosis method for an anti-recoil device of an artillery based on RMSD-DS, belonging to the field of artillery fault diagnosis. Firstly, determining a typical fault mode and fault characteristic signals of an anti-recoil device of an artillery, and acquiring probability distribution of corresponding evidences of each fault characteristic signal through a Gaussian model; then, quantitatively describing the importance degree of each evidence in the fusion decision Guo general by constructing an RMSD similarity coefficient and solving the reliability of each evidence, and accordingly, distributing weights to the evidences so as to eliminate the conflict influence among information; and finally, solving the integration evidence after weighted averaging, and applying a DS fusion rule to perform self fusion on the integration evidence to obtain a final fusion result so as to realize fault diagnosis of the gun recoil prevention device. The method for solving the basic probability distribution value of the evidence corresponding to each fault characteristic signal is simple and easy, and the fault diagnosis effect is excellent. The invention can improve the efficiency and the precision of fault diagnosis of the gun recoil prevention device when the conflict information is fused.
Description
Technical Field
The invention relates to a fault diagnosis method for an anti-recoil device of an artillery based on RMSD-DS, belonging to the technical field of artillery fault diagnosis.
Background
The anti-recoil device of the artillery is used as a key part of the artillery, and plays roles of dissipating and storing recoil energy when the artillery shoots and resetting a gun body. In a battlefield, the complexity and uncertainty of the operation environment cause the failure of the anti-recoil device of the artillery to frequently occur, if the failure of the anti-recoil device of the artillery is not diagnosed and solved in time, the launching efficiency and the shooting precision of the artillery shell can be seriously influenced, even precious fighters are delayed, and serious battlefield accidents are caused. Therefore, the method has important practical significance for the fault diagnosis method research of the anti-recoil device of the artillery.
The artillery recoil prevention device is usually used for carrying out fault diagnosis work based on multi-source information fusion on the basis of four signal sources, namely the maximum recoil displacement, the maximum recoil speed and the recoil-to-place speed. Due to the complexity of the battlefield environment, the data sensor of the anti-recoil device of the artillery is easy to be damaged or the data acquisition process is interfered, so that mutual conflict exists between the information output by the sensor. At the moment, if a traditional multi-source information fusion method is applied, similar to a neural network method, the conditions of low diagnosis efficiency and low diagnosis precision are easy to occur when the fault diagnosis mode of the gun recoil prevention device is identified, and the fault diagnosis requirement of the gun recoil prevention device cannot be met.
Disclosure of Invention
In order to solve the problems of low efficiency, low precision and the like of the conventional fault diagnosis method for the anti-recoil device of the cannon during the fusion of conflict information, the invention mainly aims to provide the fault diagnosis method for the anti-recoil device of the cannon based on RMSD-DS, and the method determines the typical fault mode and the fault characteristic signal of the anti-recoil device of the cannon; acquiring and obtaining corresponding fault characteristic signals of the gun recoil device under a typical fault mode; acquiring corresponding fault characteristic signals in a typical fault mode, and classifying the fault characteristic signals into fault training sample data and fault to-be-detected sample data; solving the average value and the standard deviation of training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals, and then constructing Gaussian models of the training samples belonging to different fault modes on different fault signals; according to the constructed failure mode Gaussian model of the gun recoil device, solving basic probability distribution of evidences corresponding to failure characteristic signals of a sample to be detected of the gun recoil device; according to the basic probability distribution of the evidence corresponding to the characteristic signal of the to-be-detected sample of the gun recoil device obtained through solving, under the framework of the failure mode of the gun recoil device, defining and solving conflict factors among the evidences, and constructing a conflict factor matrix according to all the solved conflict factors; defining and solving Root Mean Square Deviation (RMSD) distances among all evidences under a failure mode framework of the gun anti-recoil device; constructing an RMSD collision coefficient by taking the collision factor and the geometric mean value of the normalized RMSD distance as the value of the RMSD collision coefficient; according to the constructed RMSD conflict coefficient, solving and constructing an RMSD similarity coefficient; defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition; analyzing the reliability of each evidence, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence; carrying out weight distribution on all evidences according to the reliability of each evidence to reduce the conflict among information, and then obtaining an integrated evidence through weighted average; fusing the integration evidence by using a Dempster-Shafer (DS) evidence theory method under a gun fault mode frame to obtain the occurrence probability of the fault mode of the corresponding gun anti-recoil device; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing high-precision efficiency diagnosis of the gun recoil prevention device fault based on RMSD-DS.
The purpose of the invention is realized by the following technical scheme.
The invention discloses a fault diagnosis method for an anti-recoil device of an artillery based on RMSD-DS, which comprises the following steps:
the method comprises the following steps: and determining a typical fault mode and a fault characteristic signal of the anti-recoil device of the artillery.
Step two: and acquiring and obtaining corresponding fault characteristic signals of the gun recoil device under a typical fault mode.
Step three: analyzing the corresponding fault characteristic signals obtained in the step two under the typical fault mode, and classifying the fault characteristic signals into fault training sample data and fault to-be-detected sample data; solving the average value and the standard deviation of training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals, and then constructing Gaussian models of the training samples belonging to different fault modes on different fault signals.
Step four: and solving the basic probability distribution of the evidence corresponding to the fault characteristic signal of the sample to be detected of the artillery anti-recoil device according to the failure mode Gaussian model of the artillery anti-recoil device constructed in the third step.
Step five: according to the basic probability distribution of the evidence corresponding to the fault characteristic signals of the to-be-detected sample of the gun recoil device obtained by solving in the fourth step, defining and solving conflict factors among the evidences under the fault mode framework of the gun recoil device, and constructing a conflict factor matrix according to all the solved conflict factors; under the fault mode framework of the gun recoil device, RMSD distances among all evidences are defined and solved; constructing an RMSD collision coefficient by taking the collision factor and the geometric mean value of the normalized RMSD distance as the value of the RMSD collision coefficient; and solving and constructing the RMSD similarity coefficient according to the constructed RMSD collision coefficient, and facilitating the definition of the reliability of the subsequent step six through the constructed RMSD similarity coefficient.
Step six: defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition; analyzing the reliability of each evidence, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence; and performing weight distribution on all evidences according to the reliability of each evidence to reduce the conflict between information, and then obtaining an integrated evidence through weighted average, so that the integration in the subsequent step seven is facilitated, and the fault diagnosis accuracy is improved.
Step seven: in a gun fault mode frame, self-fusing the integration evidence obtained in the step six by using a Dempster-Shafer (DS) evidence theory method to obtain the occurrence probability of the fault mode of the corresponding gun anti-recoil device; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing high-precision efficiency diagnosis of the gun recoil prevention device fault based on RMSD-DS.
Further comprises the following steps: substituting the sample data to be detected of the failure of the gun recoil device determined in the third step into the failure mode Gaussian model constructed in the fourth step, and solving the basic probability distribution of the evidence corresponding to each failure characteristic signal; and (4) redistributing the weight to each evidence by using the reliability determined in the step six, reducing the influence caused by conflict information, and improving the fault diagnosis performance of the anti-recoil device. The improvement of the diagnosis performance of the anti-recoil device of the artillery comprises the improvement of the diagnosis efficiency and the diagnosis precision of the anti-recoil device.
The method comprises the following steps: and determining a typical fault mode and a fault characteristic signal of the anti-recoil device of the artillery.
Determining three typical failure modes of the artillery anti-recoil device, namely a check ring abrasion X, a re-advancing machine air leakage Y and a check rod piston abrasion Z, wherein the three typical failure modes are the check ring abrasion X, the re-advancing machine air leakage Y and the check rod piston abrasion Z, and then a failure mode frame of the artillery anti-recoil device is expressed as theta ═ X, Y and Z }; determining fault characteristic signals of the anti-recoil device of the artillery as follows: a maximum squat displacement Xmax, a maximum squat speed Vmax, a maximum reentry speed Umax, and a reentry to position speed Uend.
Step two: and acquiring and obtaining corresponding fault characteristic signals of the gun recoil device under a typical fault mode.
When the artillery works, four fault characteristic signals of the anti-recoil device in each fault mode are collected by a sensor arranged on the anti-recoil device, and the obtained data is F i The method comprises the following steps of (1) representing three fault modes of check ring wear, double-feed machine air leakage and check rod piston wear, wherein F is X, Y and Z; i is 1,2,3 and 4, which respectively represent four fault characteristic signals of a maximum recoil displacement Xmax, a maximum recoil speed Vmax, a maximum recoil speed Umax and a recoil to reach speed Uend; a set of sample data collected is denoted as (F) 1 ,F 2 ,F 3 ,F 4 )，F 1 The most corresponding to the representative failure mode FBig recoil displacement Xmax signal data, F 2 Represents the maximum squat speed Vmax signal data corresponding to the failure mode F, F 3 Representing maximum remade speed Umax signal data, F, corresponding to the failure mode F 4 Representing the signal data of the speed Uend of the double-entry bit corresponding to the failure mode F.
Step three: analyzing the corresponding fault characteristic signals obtained in the step two under the typical fault mode, and classifying the fault characteristic signals into fault training sample data and fault to-be-detected sample data; solving the average value and the standard deviation of training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals, and then constructing Gaussian models of the training samples belonging to different fault modes on different fault signals.
Step 3.1: and classifying the fault characteristic signals acquired in the step two and corresponding to the typical fault mode into fault training sample data and fault to-be-detected sample data.
And on the basis of the fault characteristic signals corresponding to the typical fault modes acquired in the step two, sample data in a preset proportion is selected from four kinds of corresponding fault characteristic signal data in each fault mode to serve as fault training samples, and the residual sample data serves as a fault to-be-detected sample.
Step 3.2: and (4) solving the average value and the standard deviation of the training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals for the fault training samples selected in the step (3.1).
For the selected fault training samples, solving the average value mu (F) of the training samples belonging to different fault modes on different fault characteristic signals i ) And standard deviation σ (F) i ) Average value of μ (F) i ) The formula (1) shows the standard deviation sigma (F) i ) Is expressed by equation (2):
in the formulas (1) and (2), F ═ X, Y, Z represent three failure modes; i is 1,2,3,4, which represents four fault characteristic signals; j ═ 1,2, …, N, and represents a data sequence;
step 3.3: and (3) constructing Gaussian models of the training samples belonging to different fault modes on different fault signals according to the average value and the standard deviation obtained by solving in the step 3.2.
Average value mu (F) obtained by solving according to step 3.2 i ) And standard deviation σ (F) i ) And constructing Gaussian models (membership function) of the training samples belonging to different fault modes on different fault signals, wherein the Gaussian models are shown in a formula (3):
in formula (3), F ═ X, Y, Z, represents three failure modes; i is 1,2,3,4, which represents four fault signatures.
When the fault characteristic signal is the maximum recoil displacement (Xmax), the fault modes are Gaussian models on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z), and the Gaussian models are as follows:
when the fault characteristic signal is the maximum recoil speed (Vmax), the fault mode is a Gaussian model on the check ring abrasion (X), the double-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is the maximum re-advancing speed (Umax), the fault modes are Gaussian models on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is a re-advancing in-place speed (Uend), the fault mode is a Gaussian model on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z), and the Gaussian model is as follows:
and
equations (4) to (15) are gaussian models of the training samples of different failure modes on different failure signals.
Step four: and solving basic probability distribution of evidences corresponding to fault characteristic signals of the samples to be detected of the artillery anti-recoil device according to the fault mode Gaussian model of the artillery anti-recoil device constructed in the third step, so that the fault diagnosis efficiency can be improved on the basis of ensuring the fault diagnosis precision of the anti-recoil device in the subsequent step.
Step 4.1: and 3.3, solving the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models with different fault modes under each fault signal according to the Gaussian models with the fault modes of the gun recoil prevention device constructed in the step 3.3.
For a set of suspect samples for which the failure mode is unknown, the corresponding data can be represented as (F) 1 ,F 2 ,F 3 ,F 4 ) Wherein F is X, Y, Z; the subscripts have a value of 1 for the maximum squat displacement (Xmax), 2 for the maximum squat speed (Vmax), 3 for the maximum re-travel speed (Umax), and 4 for the re-travel to reach speed (Uend).
When the fault characteristic signal is maximum recoil displacement (Xmax), the solution formula of the ordinate of the intersection point of the sample to be detected and the Gaussian model of different fault modes is as follows:
when the fault characteristic signal is the maximum recoil speed (Vmax), the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
when the fault characteristic signal is the maximum recurrence velocity (Umax), the ordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
when the fault characteristic signal is a complex-in-place speed (Uend), the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
step 4.2: and the basic probability distribution of the evidences corresponding to the four fault characteristic signals is shown.
Each sample to be checked contains four fault characteristic signals (Xmax, Vmax, Umax and Uend), each fault characteristic signal corresponds to one group of evidences, and then the basic probability distribution function of the evidences corresponding to each fault characteristic signal (Xmax, Vmax, Umax and Uend) can be represented as m i (i ═ 1,2,3, 4); in the step one, the failure modes of the gun recoil preventing device are X, Y and Z, and the basic probability distribution of a group of evidences comprises m i (X)、m i (Y) and m i (Z) wherein m i (X) is shown in the evidence m i The lower to-be-examined sample belongs to the fundamental probability distribution, m, of the failure mode X i (Y) is shown in the evidence m i The lower to-be-examined sample belongs to the fundamental probability distribution of the failure mode Y, m i (Z) is shown in the evidence m i The lower suspect sample belongs to the fundamental probability distribution of the failure mode Z.
Step 4.3: and 4.1, acquiring the vertical coordinate of the intersection point of the to-be-detected sample and the Gaussian models with different fault modes under each fault signal, and respectively solving the basic probability distribution of the evidence corresponding to each fault characteristic signal of the to-be-detected sample.
Evidence m corresponding to the fault characteristic signal of the maximum recoil displacement (Xmax) 1 The solving formula of the basic probability distribution function is as follows:
corresponding evidence m when the fault signature is at maximum squat speed (Vmax) 2 The basic probability distribution function of (a) is: m is 2 (X)，m 2 (Y)，m 2 (Z)；
Corresponding evidence m when the fault characteristic signal is the maximum re-entry speed (Umax) 3 The basic probability distribution function of (a) is: m is 3 (X)，m 3 (Y)，m 3 (Z)；
The corresponding evidence m when the fault characteristic signal is the speed of the multiple entry (Uend) 4 The basic probability distribution function of (a) is: m is 4 (X)，m 4 (Y)，m 4 (Z)。
Step five: according to the basic probability distribution of the evidence corresponding to the fault characteristic signals of the to-be-detected sample of the gun recoil device obtained by solving in the fourth step, defining and solving conflict factors among the evidences under the fault mode framework of the gun recoil device, and constructing a conflict factor matrix according to all the solved conflict factors; under the fault mode framework of the gun recoil device, RMSD distances among all evidences are defined and solved; constructing an RMSD collision coefficient by taking the collision factor and the geometric mean value of the normalized RMSD distance as the value of the RMSD collision coefficient; and solving and constructing the RMSD similarity coefficient according to the constructed RMSD collision coefficient, and facilitating the definition of the reliability of the subsequent step six through the constructed RMSD similarity coefficient.
Step 5.1: and defining and solving conflict factors among the evidences under a failure mode framework of the gun recoil prevention device, and constructing a conflict factor matrix according to all the solved conflict factors. The conflict factor between the various evidences is the conflict factor between each two sets of evidences.
To conveniently provide a formula for solving the conflict factor, m is defined under the framework of artillery fault modes theta ═ X, Y and Z- 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ″ ═ X, Y, Z), respectively, and evidence m is evidence 1 And m 2 Is as shown in equation (40):
the collision factor matrix among the evidences according to the formula (40) is:
however, the collision factor has a defect, and it can be known from the formula (40) that the collision factor obtained by solving when the two evidences are the same is not 0, so that the collision factor needs to be corrected by introducing the RMSD distance between the evidences in the subsequent step 5.2, and the collision coefficient, i.e., the RMSD collision coefficient, is constructed in the subsequent step 5.3.
Step 5.2: under the fault mode framework of the gun recoil device, RMSD distances among all evidences are defined and solved, an RMSD distance matrix is constructed according to all the solved RMSD distances, and the normalized RMSD distance matrix is solved.
Definition m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ═ X, Y, Z), respectively, and evidence m is defined 1 And m 2 The Root Mean Square Deviation (RMSD) distance between is:
and (3) solving according to a formula (42) to obtain an RMSD distance matrix among the evidences, wherein the formula (43) is as follows:
then searching the maximum value RMSD in the RMSD distance matrix max Max { RMSD }, followed by normalization of the RMSD distance matrix, i.e., each element of the RMSD distance matrix divided by the RMSD max And obtaining a normalized RMSD distance matrix, as shown in formula (44):
step 5.3: and defining the collision factor in the step 5.1 and the geometric mean value of the normalized RMSD distance in the step 5.2 as the value of the RMSD collision coefficient, and constructing the RMSD collision coefficient.
Defining evidence m by taking the collision factor obtained in the step 5.1 and the geometric mean of the normalized RMSD distance obtained in the step 5.2 as the value of the RMSD collision coefficient 1 And m 2 The RMSD collision coefficient of (a) is expressed as:
in formula (45), K (m) 1 ,m 2 ) Represents evidence m 1 And m 2 Conflict factor of (2), RMSD (m) 1 ,m 2 ) Represents evidence m 1 And m 2 Normalized RMSD distance;
the RMSD collision coefficient matrix obtained according to equation (45) is:
step 5.4: the RMSD similarity coefficient is solved and constructed based on the RMSD collision coefficient described in step 5.3.
The RMSD collision coefficient constructed in the step 5.3 represents the collision degree among evidences, and the value range is [0,1 ]]Subtracting the RMSD collision coefficient from 1Defining the evidence m to represent the similarity degree between the evidences 1 And m 2 The RMSD similarity coefficient of (a) is expressed as:
Sim RMSD (m 1 ,m 2 )＝1-Con RMSD (m 1 ,m 2 ) (47)
the RMSD similarity coefficient matrix obtained according to equation (47) is:
step six: defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition; analyzing the reliability of each evidence, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence; and performing weight distribution on all evidences according to the reliability of each evidence to reduce the conflict between information, and then obtaining an integrated evidence through weighted average, so that the integration in the subsequent step seven is facilitated, and the fault diagnosis accuracy is improved.
Step 6.1: defining the reliability of each evidence as the sum of the RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition.
Defining the reliability of each evidence as the sum of the RMSD similarity coefficients of the evidence and other evidences, and determining the evidence m 1 The reliability of (d) is formulated as:
Rel(m 1 )＝Sim RMSD (m 1 ,m 2 )+Sim RMSD (m 1 ,m 3 )+Sim RMSD (m 1 ,m 4 ) (49)
evidence m 2 Is formulated as:
Rel(m 2 )＝Sim RMSD (m 2 ,m 1 )+Sim RMSD (m 2 ,m 3 )+Sim RMSD (m 2 ,m 4 ) (50)
evidence m 3 The reliability of (d) is formulated as:
Rel(m 3 )＝Sim RMSD (m 3 ,m 1 )+Sim RMSD (m 3 ,m 2 )+Sim RMSD (m 3 ,m 4 ) (51)
evidence m 4 The reliability of (d) is formulated as:
Rel(m 4 )＝Sim RMSD (m 4 ,m 1 )+Sim RMSD (m 4 ,m 2 )+Sim RMSD (m 4 ,m 3 ) (52)
wherein, the reliability of the evidence represents the support degree of other evidence to the evidence;
the greater the reliability of the evidence is, the higher the importance of the evidence in the fusion decision process is, and the weight distributed in the subsequent step 6.2 is preferably large;
the smaller the reliability of the evidence is, the lower the importance of the evidence in the fusion decision process is, and the smaller the weight assigned in the subsequent step 6.2 is.
Step 6.2: the reliability of each evidence determined in step 6.1 is analyzed, the weight of each evidence is defined as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and the weight of each evidence is solved.
Defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of all evidence reliabilities, and determining the evidence m 1 The weight of (d) is expressed as:
evidence m 2 The reliability of (d) is formulated as:
evidence m 3 The reliability of (d) is formulated as:
evidence m 4 The reliability of (d) is formulated as:
step 6.3: and (4) performing weight distribution on all evidences according to the reliability of each evidence in the step 6.2, reducing the conflict between information, and then obtaining an integrated evidence through weighted average, so that the integration in the subsequent step seven is facilitated, and the fault diagnosis accuracy is improved.
The integrated evidence is obtained by weighted averaging according to the weights assigned in step 6.2, and is represented as:
further, a basic probability distribution is obtained for failure modes of X, Y and Z, respectively, under the evidence of integration, expressed as:
step seven: in a gun fault mode frame, self-fusing the integration evidence obtained in the step six by using a Dempster-Shafer (DS) evidence theory method to obtain the occurrence probability of the fault mode of the corresponding gun anti-recoil device; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing high-precision efficiency diagnosis of the gun recoil prevention device fault based on RMSD-DS.
Step 7.1: definition m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted F '(F' ═ X, Y, Z) and F ″ (F ═ X, Y, Z), respectively, giving evidence m 1 And m 2 The DS fusion rule of (1).
To give DS fusion rules for convenience, m is set under the artillery fault mode framework Θ ═ { X, Y, Z } 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ″ ═ X, Y, Z), respectively, and evidence m is evidence 1 And m 2 The DS fusion rule of (1) is shown as formula (61):
step 7.2: and fusing the integration evidence for 3 times according to the DS fusion rule given in the step 7.1, and fusing the integration evidence by using the DS fusion rule to obtain the occurrence probability of the fault mode of the corresponding artillery recoil prevention device.
Step 7.3: and (3) obtaining the occurrence probability of the fault modes of all the gun recoil resisting devices according to the traversal in the step (7.2), determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing the high-precision and efficient diagnosis of the fault of the gun recoil resisting devices based on an RMSD-DS method.
Has the advantages that:
1. compared with traditional fusion methods such as a neural network and the like, the RMSD-DS-based artillery recoil device fault diagnosis method disclosed by the invention has the advantages that under the framework of a failure mode of an artillery recoil device, the importance degree of each evidence is quantitatively described by constructing an RMSD similarity coefficient and solving the reliability of each evidence, weight distribution is carried out on all evidences according to the reliability of each evidence, the influence caused by conflict information is reduced, integrated evidences are obtained after weighted averaging, and the DS method is used for integrating the evidences to fuse, so that the occurrence probability of the failure mode of the corresponding artillery recoil device is obtained; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, and improving the fault diagnosis precision of the gun recoil prevention devices.
2. The invention discloses a fault diagnosis method for an anti-recoil device of an artillery based on RMSD-DS, which comprises the steps of establishing a Gaussian model of a fault mode of the anti-recoil device of the artillery by analyzing fault historical data of the anti-recoil device of the artillery, solving basic probability distribution of evidences corresponding to fault characteristic signals of the anti-recoil device of the artillery by the Gaussian model of the fault mode of the anti-recoil device of the artillery, and improving the fault diagnosis efficiency of the anti-recoil device of the artillery.
3. The fault diagnosis method for the gun recoil device based on the RMSD-DS disclosed by the invention realizes high-precision and efficient diagnosis of the fault of the gun recoil device based on the fault diagnosis method for the gun recoil device based on the RMSD-DS on the basis of realizing the beneficial effects 1 and 2.
Drawings
FIG. 1 is a flow chart of the fault diagnosis method of the anti-recoil device of the artillery based on RMSD-DS.
FIG. 2 is a Gaussian model of various modes under different fault signature, wherein: fig. 2(a) shows gaussian models of various failure modes when the failure characteristic signal is Xmax, fig. 2(b) shows gaussian models of various failure modes when the failure characteristic signal is Vmax, fig. 2(c) shows gaussian models of various failure modes when the failure characteristic signal is Umax, and fig. 2(d) shows gaussian models of various failure modes when the failure characteristic signal is Uend.
FIG. 3 is a flow chart for solving RMSD similarity coefficients between two evidences.
Detailed Description
For better illustrating the objects and advantages of the present invention, the following description will be made with reference to the accompanying drawings and examples, and compared with the conventional neural network fusion method and DS evidence theory method.
As shown in FIG. 1; the method for diagnosing the fault of the anti-recoil device of the artillery based on the RMSD-DS disclosed by the embodiment comprises the following specific implementation steps of:
the method comprises the following steps: and determining a typical fault mode and a fault characteristic signal of the anti-recoil device of the artillery.
Determining three typical failure modes of the anti-recoil device of the artillery, namely a check ring wear X, a re-advancing machine air leakage Y and a check rod piston wear Z, wherein the failure mode frame of the anti-recoil device of the artillery is expressed as theta (X, Y and Z); determining fault characteristic signals of the anti-recoil device of the artillery as follows: a maximum squat displacement Xmax, a maximum squat speed Vmax, a maximum reentry speed Umax, and a reentry to position speed Uend.
Step two: and acquiring and obtaining corresponding fault characteristic signals of the gun recoil device under a typical fault mode.
When the artillery works, four fault characteristic signals of the anti-recoil device in each fault mode are collected by a sensor arranged on the anti-recoil device, and the obtained data is F i The method comprises the following steps of (1) representing three fault modes of check ring wear, double-feed machine air leakage and check rod piston wear, wherein F is X, Y and Z; i is 1,2,3 and 4, which respectively represent four fault characteristic signals of a maximum recoil displacement Xmax, a maximum recoil speed Vmax, a maximum recoil speed Umax and a recoil to reach speed Uend; a set of sample data collected is denoted as (F) 1 ,F 2 ,F 3 ,F 4 ) I.e. F 1 Representing maximum recoil displacement Xmax signal data, F, corresponding to failure mode F 2 Represents the maximum squat speed Vmax signal data corresponding to the failure mode F, F 3 Representing maximum remade speed Umax signal data, F, corresponding to the failure mode F 4 Representing the signal data of the speed Uend of the double-entry bit corresponding to the failure mode F.
In each failure mode, 100 sets of failure data were obtained for a total of 300 sets of data.
Step three: analyzing the corresponding fault characteristic signals obtained in the step two under the typical fault mode, and classifying the fault characteristic signals into fault training sample data and fault to-be-detected sample data; solving the average value and the standard deviation of training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals, and then constructing Gaussian models of the training samples belonging to different fault modes on different fault signals.
Step 3.1: and classifying the fault characteristic signals acquired in the step two and corresponding to the typical fault mode into fault training sample data and fault to-be-detected sample data.
Based on the fault characteristic signals corresponding to the typical fault mode obtained in the second step, 80% of sample data is selected from four kinds of corresponding fault characteristic signal data in each fault mode to be used as fault training samples, and the remaining 20% of sample data is used as a fault to be detected sample.
In order to verify the effectiveness of the method, the damage of the artillery anti-recoil device sensor is simulated by exchanging the maximum recoil displacement data corresponding to the fault mode Y and the fault mode Z; the purpose of this process is to allow the information output by the sensors to conflict with each other.
Step 3.2: and (4) solving the average value and the standard deviation of the training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals for the fault training samples selected in the step (3.1).
The results of the solution are shown in table 1.
TABLE 1 mean and standard deviation of training samples (different failure modes)
Step 3.3: and (3) constructing Gaussian models of the training samples belonging to different fault modes on different fault signals according to the average value and the standard deviation obtained by solving in the step 3.2.
When the fault characteristic signal is the maximum recoil displacement (Xmax), the fault modes are Gaussian models of the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z), and the Gaussian models are as follows:
when the fault characteristic signal is the maximum recoil speed (Vmax), the fault mode is a Gaussian model on the check ring abrasion (X), the double-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is the maximum re-advancing speed (Umax), the fault modes are Gaussian models on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is a re-advancing in-place speed (Uend), the fault modes are Gaussian models of the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z), and the Gaussian models are as follows:
and
after solving the above gaussian model, under each fault characteristic signal, a gaussian model of each fault mode is drawn, as shown in fig. 2.
Step four: and solving basic probability distribution of evidences corresponding to the fault characteristic signals of the to-be-detected samples of the artillery anti-recoil device according to the Gaussian model of the fault mode of the artillery anti-recoil device constructed in the third step.
In each failure mode, 20 groups of samples to be detected are respectively provided, and due to text space limitation, only a group of basic probability distribution of corresponding evidences of each failure characteristic signal of the sample data to be detected is given from each failure mode, as shown in tables 2 to 4.
TABLE 2 basic probability distribution values of corresponding evidences for each fault characteristic signal when the actual fault is X
For the set of samples to be examined, it can be seen from Table 2 that when the actual failure is X, the evidence m is 1 、m 2 And m 3 Supporting failure mode X occurrence, evidence m 4 Supporting failure mode Y occurs, there is a conflict situation between the evidences.
TABLE 3 basic probability distribution value of evidence corresponding to each fault characteristic signal when the actual fault is Y
For this set of samples to be examined, it can be seen from Table 3 that when the actual failure is Y, the evidence m 2 、m 3 And m 4 Supporting failure mode Y occurrence, evidence m 1 Supporting failure mode Z to occur, there is a conflict situation between the evidences.
TABLE 4 basic probability distribution values of evidence corresponding to each fault characteristic signal when the actual fault is Z
For this set of samples to be examined, it can be seen from Table 4 that when the actual failure is Z, the evidence m is 1 Supporting failure mode X occurrence, evidence m 2 、m 3 And m 4 Supporting failure mode Z to occur, there is a conflict situation between the evidences.
Step five: according to the basic probability distribution of the evidence corresponding to the fault characteristic signals of the to-be-detected sample of the gun recoil device obtained by solving in the fourth step, defining and solving conflict factors among the evidences under the fault mode framework of the gun recoil device, and constructing a conflict factor matrix according to all the solved conflict factors; under the fault mode framework of the gun recoil device, RMSD distances among all evidences are defined and solved; constructing an RMSD collision coefficient by taking the collision factor and the geometric mean value of the normalized RMSD distance as the value of the RMSD collision coefficient; and solving and constructing the RMSD similar coefficient according to the constructed RMSD conflict coefficient, and conveniently defining the reliability of the subsequent step six through the constructed RMSD similar coefficient.
A flow chart for solving the RMSD similarity coefficient between two pieces of evidence is shown in fig. 3.
Based on the data given in the step four, the RMSD similarity coefficient obtained by solving is as follows,
when the actual fault is X, the RMSD similarity coefficient matrix among the evidences of a group of sample data to be detected is as follows:
when the actual fault is Y, the RMSD similarity coefficient matrix among the evidences of a group of sample data to be detected is as follows:
when the actual fault is Z, the RMSD similarity coefficient matrix among the evidences of a group of sample data to be detected is as follows:
step six: defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition; analyzing the reliability of each evidence, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence; and performing weight distribution on all evidences according to the reliability of each evidence to reduce the conflict between information, and then obtaining an integrated evidence through weighted average, so that the integration in the subsequent step seven is facilitated, and the fault diagnosis accuracy is improved.
When the actual fault is X, the integrated evidence of one group of sample data to be detected is as follows:
when the actual fault is Y, the integrated evidence of one group of sample data to be detected is as follows:
when the actual fault is Z, the integrated evidence of one group of sample data to be detected is as follows:
step seven: the DS of the integrated evidence is fused with the output of the diagnostic result.
And fusing the integration evidence for 3 times according to the DS fusion rule to obtain a final fusion result.
When the actual fault is X, the final fusion result of a group of sample data to be detected is as follows:
m(X)＝0.9999，m(Y)＝3.44e-05，m(Z)＝4.83e-05；
the final diagnostic mode is: x; the diagnosis result is correct.
When the actual fault is Y, the final fusion result of one group of the sample data to be detected is as follows:
m(X)＝0.0081，m(Y)＝0.9886，m(Z)＝0.0033；
the final diagnostic mode is: y; the diagnosis result is correct.
When the actual fault is Z, the final fusion result of a group of sample data to be detected is as follows:
m(X)＝9.16e-04，m(Y)＝1.27e-05，m(Z)＝0.9991；
the final diagnostic mode is: z; the diagnosis result is correct.
According to the same method, the diagnosis results of all samples to be examined are obtained by solving, as shown in Table 5.
TABLE 5 diagnosis results of all samples to be examined (method of this patent)
The results in table 5 show that the method provided by the patent has the fault diagnosis accuracy of 100% for the fault modes X and Y, 95% for the fault mode Z, and 98.3% for the total fault diagnosis accuracy, which indicates that the method has outstanding diagnosis effect and excellent diagnosis accuracy.
Step seven: in a gun fault mode frame, self-fusing the integration evidence obtained in the step six by using a Dempster-Shafer (DS) evidence theory method to obtain the occurrence probability of the fault mode of the corresponding gun anti-recoil device; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing high-precision efficiency diagnosis of the gun recoil prevention device fault based on RMSD-DS.
In order to further highlight the diagnostic effect of the method of the present patent, the diagnostic results obtained by applying the DS fusion rule and the BP neural network method are given, as shown in tables 6 to 7.
TABLE 6 diagnosis results (DS) of all samples to be examined
The results in table 6 show that the DS method has a failure diagnosis accuracy of 100% for the failure mode X, 1% for the failure mode Y, 95% for the failure mode Y, and 68.3% for the total failure diagnosis accuracy, which is far lower than the diagnosis accuracy of the method provided in this patent.
TABLE 7 diagnosis results (BP neural network) of all samples to be examined
The results in table 7 show that the BP neural network method has a failure diagnosis accuracy of 70% for the failure mode X, 100% for the failure mode Y, 75% for the failure mode Y, and 81.67% for the total failure diagnosis accuracy, which is far lower than the diagnosis accuracy of the method provided by the present patent.
As can be seen from a combination of tables 6 and 7, the present example is superior in diagnostic effect and has higher diagnostic efficiency and accuracy.
Step eight: substituting the sample data to be detected of the failure of the gun recoil device determined in the third step into the failure mode Gaussian model constructed in the fourth step, and solving the basic probability distribution of the evidence corresponding to each failure characteristic signal; and (4) redistributing the weight to each evidence by using the reliability determined in the step six, reducing the influence caused by conflict information, and improving the fault diagnosis performance of the anti-recoil device. The improvement of the diagnosis performance of the anti-recoil device of the artillery comprises the improvement of the diagnosis efficiency and the diagnosis precision of the anti-recoil device.
The above detailed description of the present invention further illustrates the objects, technical solutions and effects of the present invention, but the embodiments of the present invention are not limited thereto, and any modifications made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (9)
1. The RMSD-DS-based gun recoil device fault diagnosis method is characterized by comprising the following steps of:
the method comprises the following steps: determining a typical fault mode and a fault characteristic signal of the gun recoil preventing device;
step two: acquiring and obtaining corresponding fault characteristic signals of the gun recoil device under a typical fault mode;
step three: analyzing the corresponding fault characteristic signals obtained in the step two under the typical fault mode, and classifying the fault characteristic signal data into fault training sample data and fault to-be-detected sample data; solving the average value and the standard deviation of training samples belonging to different fault modes of the gun recoil resisting device on different fault characteristic signals, and then constructing Gaussian models of the training samples belonging to different fault modes on different fault signals;
step four: solving basic probability distribution of evidences corresponding to fault characteristic signals of the to-be-detected sample of the artillery anti-recoil device according to the failure mode Gaussian model of the artillery anti-recoil device constructed in the third step;
step five: according to the basic probability distribution of the evidence corresponding to the fault characteristic signals of the to-be-detected sample of the gun recoil device obtained by solving in the fourth step, defining and solving conflict factors among the evidences under the fault mode framework of the gun recoil device, and constructing a conflict factor matrix according to all the solved conflict factors; defining and solving Root Mean Square Deviation (RMSD) distances among all evidences under a failure mode framework of the gun anti-recoil device; constructing an RMSD collision coefficient by taking the collision factor and the geometric mean value of the normalized RMSD distance as the value of the RMSD collision coefficient; solving and constructing an RMSD similarity coefficient according to the constructed RMSD conflict coefficient, and conveniently defining the reliability of the subsequent step six through the constructed RMSD similarity coefficient;
step six: defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition; analyzing the reliability of each evidence, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence; carrying out weight distribution on all evidences according to the reliability of each evidence to reduce the conflict between information, and then obtaining an integrated evidence through weighted average, thereby facilitating the fusion of the subsequent step seven and improving the accuracy of fault diagnosis;
step seven: in a gun fault mode frame, self-fusing the integration evidence obtained in the step six by using a Dempster-Shafer (DS) evidence theory method to obtain the occurrence probability of the fault mode of the corresponding gun anti-recoil device; and traversing the occurrence probability of the fault modes of all the gun recoil prevention devices, determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing high-precision efficiency diagnosis of the gun recoil prevention device fault based on RMSD-DS.
2. The RMSD-DS based gun recoil device fault diagnosis method of claim 1, further comprising the steps of: substituting the sample data to be detected of the failure of the gun recoil device determined in the third step into the failure mode Gaussian model constructed in the fourth step, and solving the basic probability distribution of the evidence corresponding to each failure characteristic signal; the reliability determined in the step six is used for redistributing the weight to each evidence, so that the influence caused by conflict information is reduced, and the fault diagnosis performance of the anti-recoil device is improved; the improvement of the diagnosis performance of the anti-recoil device comprises improvement of diagnosis efficiency and diagnosis precision of the anti-recoil device.
3. The RMSD-DS-based gun recoil device fault diagnosis method as claimed in claim 1 or 2, characterized in that the first step is realized by the method as follows:
determining three typical failure modes of the artillery anti-recoil device, namely a check ring abrasion X, a re-advancing machine air leakage Y and a check rod piston abrasion Z, wherein the three typical failure modes are the check ring abrasion X, the re-advancing machine air leakage Y and the check rod piston abrasion Z, and then a failure mode frame of the artillery anti-recoil device is expressed as theta ═ X, Y and Z }; the fault characteristic signals of the gun anti-recoil device are determined as follows: a maximum squat displacement Xmax, a maximum squat speed Vmax, a maximum reentry speed Umax, and a reentry to position speed Uend.
4. The RMSD-DS-based gun recoil device fault diagnosis method as claimed in claim 3, wherein the second step is realized by the following method:
when the artillery works, four fault characteristic signals of the anti-recoil device in each fault mode are collected by a sensor arranged on the anti-recoil device, and the obtained data is F i The method comprises the following steps of (1) representing three fault modes of check ring wear, double-feed machine air leakage and check rod piston wear, wherein F is X, Y and Z; i is 1,2,3 and 4, which respectively represent four fault characteristic signals of a maximum recoil displacement Xmax, a maximum recoil speed Vmax, a maximum recoil speed Umax and a recoil to reach speed Uend; a set of collected sample dataIs represented by (F) 1 ,F 2 ,F 3 ,F 4 )，F 1 Representing maximum recoil displacement Xmax signal data, F, corresponding to failure mode F 2 Represents the maximum squat speed Vmax signal data corresponding to the failure mode F, F 3 Representing maximum remade speed Umax signal data, F, corresponding to the failure mode F 4 Representing the signal data of the speed Uend of the double-entry bit corresponding to the failure mode F.
5. The RMSD-DS based artillery recoil device fault diagnosis method of claim 4, wherein the third step comprises the following steps:
step 3.1: classifying the fault characteristic signals corresponding to the typical fault mode acquired in the step two into fault training sample data and fault to-be-detected sample data;
based on the fault characteristic signals corresponding to the typical fault modes acquired in the step two, sample data in a preset proportion is selected from four corresponding fault characteristic signal data in each fault mode to serve as fault training samples, and the residual sample data serves as a fault to-be-detected sample;
step 3.2: for the fault training samples selected in the step 3.1, solving the average value and the standard deviation of the training samples belonging to different fault modes of the gun recoil device on different fault characteristic signals;
for the selected fault training samples, solving the average value mu (F) of the training samples belonging to different fault modes on different fault characteristic signals i ) And standard deviation σ (F) i ) Average value of μ (F) i ) The formula (1) shows the standard deviation sigma (F) i ) Is expressed by equation (2):
in the formulas (1) and (2), F ═ X, Y, Z represent three failure modes; i is 1,2,3,4, which represents four fault characteristic signals; j ═ 1,2, …, N, and represents a data sequence;
step 3.3: constructing Gaussian models of training samples belonging to different fault modes on different fault signals according to the average value and the standard deviation obtained by solving in the step 3.2;
average value mu (F) obtained by solving according to step 3.2 i ) And standard deviation σ (F) i ) And constructing Gaussian models (membership function) of the training samples belonging to different fault modes on different fault signals, wherein the Gaussian models are shown in a formula (3):
in formula (3), F ═ X, Y, Z, represents three failure modes; i is 1,2,3,4, which represents four fault characteristic signals;
when the fault characteristic signal is the maximum recoil displacement (Xmax), the fault modes are Gaussian models on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z), and the Gaussian models are as follows:
when the fault characteristic signal is the maximum recoil speed (Vmax), the fault mode is a Gaussian model on the check ring abrasion (X), the double-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is the maximum re-advancing speed (Umax), the fault modes are Gaussian models on the check ring abrasion (X), the re-advancing machine air leakage (Y) and the check rod piston abrasion (Z) as follows:
and
when the fault characteristic signal is the speed of the double-in-place (Uend), the fault mode is the controlThe Gaussian models of the ring abrasion (X), the air leakage (Y) of the re-advancing machine and the piston abrasion (Z) of the braking and retreating rod are as follows:
and
equations (4) to (15) are gaussian models of the training samples of different failure modes on different failure signals.
6. The RMSD-DS based artillery recoil device failure diagnostic method of claim 5, wherein the fourth step comprises the steps of:
step 4.1: solving the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models with different fault modes under each fault signal according to the Gaussian models with the fault modes of the gun recoil device constructed in the step 3.3;
for a set of suspect samples for which the failure mode is unknown, the corresponding data can be represented as (F) 1 ,F 2 ,F 3 ,F 4 ) Wherein F is X, Y, Z; 1 in the subscripts represents the maximum squat displacement (Xmax), 2 represents the maximum squat speed (Vmax), 3 represents the maximum re-entry speed (Umax), and 4 represents the re-entry to position speed (Uend);
when the fault characteristic signal is the maximum recoil displacement (Xmax), the solving formula of the vertical coordinate of the intersection point of the sample to be detected and the Gaussian model of different fault modes is as follows:
when the fault characteristic signal is the maximum recoil speed (Vmax), the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
when the fault characteristic signal is the maximum recurrence velocity (Umax), the ordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
when the fault characteristic signal is a complex-in-place speed (Uend), the vertical coordinate of the intersection point of the sample to be detected and the Gaussian models of different fault modes is as follows:
step 4.2: representing basic probability distribution of evidences corresponding to the four fault characteristic signals;
each sample to be checked contains four fault characteristic signals (Xmax, Vmax, Umax and Uend), each fault characteristic signal corresponds to one group of evidences, and then the basic probability distribution function of the evidences corresponding to each fault characteristic signal (Xmax, Vmax, Umax and Uend) can be represented as m i (i ═ 1,2,3, 4); in the step one, the failure modes of the gun recoil preventing device are X, Y and Z, and the basic probability distribution of a group of evidences comprises m i (X)、m i (Y) and m i (Z) wherein m i (X) is shown in the evidence m i Fundamental probability of a lower inspected sample belonging to failure mode XDistribution, m i (Y) is shown in the evidence m i The lower to-be-examined sample belongs to the fundamental probability distribution of the failure mode Y, m i (Z) is shown in the evidence m i The lower sample to be detected belongs to the basic probability distribution of the fault mode Z;
step 4.3: according to the step 4.1, acquiring the vertical coordinate of the intersection point of the to-be-detected sample and the Gaussian models with different fault modes under each fault signal, and respectively solving the basic probability distribution of the evidence corresponding to each fault characteristic signal of the to-be-detected sample;
evidence m corresponding to the fault characteristic signal of the maximum recoil displacement (Xmax) 1 The solving formula of the basic probability distribution function is as follows:
corresponding evidence m when the fault signature is at maximum squat speed (Vmax) 2 The basic probability distribution function of (a) is: m is 2 (X)，m 2 (Y)，m 2 (Z)；
Corresponding evidence m when the fault characteristic signal is the maximum re-entry speed (Umax) 3 The basic probability distribution function of (a) is: m is 3 (X)，m 3 (Y)，m 3 (Z)；
The corresponding evidence m when the fault characteristic signal is the speed of the multiple entry (Uend) 4 The basic probability distribution function of (a) is: m is 4 (X)，m 4 (Y)，m 4 (Z)；
7. The RMSD-DS based artillery recoil device fault diagnosis method of claim 6 wherein the step five comprises the steps of:
step 5.1: defining and solving conflict factors among all evidences under a failure mode framework of the gun recoil prevention device, and constructing a conflict factor matrix according to all solved conflict factors; the conflict factors among the evidences are the conflict factors among every two groups of evidences;
to conveniently give a formula for solving the conflict factor, m is defined under a cannon fault mode framework theta ═ X, Y, Z ═ m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ″ ═ X, Y, Z), respectively, and evidence m is evidence 1 And m 2 Is as shown in equation (40):
the collision factor matrix among the evidences according to the formula (40) is:
however, the collision factor has a defect, and it can be known from the formula (40) that the collision factor obtained by solving when the two evidences are the same is not 0, so that the collision factor needs to be corrected by introducing the RMSD distance between the evidences in the subsequent step 5.2, and a collision coefficient, that is, the RMSD collision coefficient, is constructed in the subsequent step 5.3;
step 5.2: under the fault mode framework of the gun recoil device, RMSD distances among all evidences are defined and solved, an RMSD distance matrix is constructed according to all the solved RMSD distances, and the normalized RMSD distance matrix is solved;
definition m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ═ X, Y, Z), respectively, and evidence m is defined 1 And m 2 The Root Mean Square Deviation (RMSD) distance between is:
and (3) solving according to a formula (42) to obtain an RMSD distance matrix among the evidences, wherein the formula (43) is as follows:
then searching the maximum value RMSD in the RMSD distance matrix max Max { RMSD }, followed by normalization of the RMSD distance matrix, i.e., each element of the RMSD distance matrix divided by the RMSD max And obtaining a normalized RMSD distance matrix, as shown in formula (44):
step 5.3: defining the collision factor in the step 5.1 and the geometric mean value of the normalized RMSD distance in the step 5.2 as the value of the RMSD collision coefficient, and constructing the RMSD collision coefficient;
defining evidence m by taking the collision factor obtained in the step 5.1 and the geometric mean of the normalized RMSD distance obtained in the step 5.2 as the value of the RMSD collision coefficient 1 And m 2 The RMSD collision coefficient of (a) is expressed as:
in formula (45), K (m) 1 ,m 2 ) Representative evidence m 1 And m 2 Conflict factor of (2), RMSD (m) 1 ,m 2 ) Represents evidence m 1 And m 2 Normalized RMSD distance;
the RMSD collision coefficient matrix obtained according to equation (45) is:
step 5.4: based on the RMSD collision coefficient in the step 5.3, solving and constructing an RMSD similarity coefficient;
the RMSD collision coefficient constructed in step 5.3 represents the collision degree among evidences, and the value range is [0,1 ]]Subtracting the RMSD collision coefficient from 1 can represent the similarity degree between the evidences, and define the evidence m 1 And m 2 The RMSD similarity coefficient of (a) is expressed as:
Sim RMSD (m 1 ,m 2 )＝1-Con RMSD (m 1 ,m 2 ) (47)
the RMSD similarity coefficient matrix obtained according to equation (47) is:
8. the RMSD-DS based artillery recoil device failure diagnostic method of claim 7, wherein the sixth step comprises the steps of:
step 6.1: defining the reliability of each evidence as the sum of RMSD similarity coefficients of the evidence and other evidences, and determining the reliability of each evidence according to the definition;
defining the reliability of each evidence as the sum of the RMSD similarity coefficients of the evidence and other evidences, and determining the evidence m 1 The reliability of (d) is formulated as:
Rel(m 1 )＝Sim RMSD (m 1 ,m 2 )+Sim RMSD (m 1 ,m 3 )+Sim RMSD (m 1 ,m 4 ) (49)
evidence m 2 The reliability of (d) is formulated as:
Rel(m 2 )＝Sim RMSD (m 2 ,m 1 )+Sim RMSD (m 2 ,m 3 )+Sim RMSD (m 2 ,m 4 ) (50)
evidence m 3 The reliability of (d) is formulated as:
Rel(m 3 )＝Sim RMSD (m 3 ,m 1 )+Sim RMSD (m 3 ,m 2 )+Sim RMSD (m 3 ,m 4 ) (51)
evidence m 4 The reliability of (d) is formulated as:
Rel(m 4 )＝Sim RMSD (m 4 ,m 1 )+Sim RMSD (m 4 ,m 2 )+Sim RMSD (m 4 ,m 3 ) (52)
wherein, the reliability of the evidence represents the support degree of other evidence to the evidence;
the greater the reliability of the evidence is, the higher the importance of the evidence in the fusion decision process is, and the weight distributed in the subsequent step 6.2 is preferably large;
the smaller the reliability of the evidence is, the lower the importance of the evidence in the fusion decision process is, and the weight distributed in the subsequent step 6.2 is preferably small;
step 6.2: analyzing the reliability of each evidence determined in the step 6.1, defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of the reliabilities of all the evidences, and solving the weight of each evidence;
defining the weight of each evidence as the ratio of the reliability of the evidence to the sum of all evidence reliabilities, and determining the evidence m 1 The weight of (d) is expressed as:
evidence m 2 The reliability of (d) is formulated as:
evidence m 3 The reliability of (d) is formulated as:
evidence m 4 The reliability of (d) is formulated as:
step 6.3: performing weight distribution on all evidences according to the reliability of each evidence in the step 6.2, reducing the conflict between information, and then obtaining an integrated evidence through weighted average, so that the integration in the subsequent step seven is facilitated, and the fault diagnosis accuracy is improved;
the integrated evidence is obtained by weighted averaging according to the weights assigned in step 6.2, and is represented as:
the basic probability distribution for failure modes of X, Y and Z, respectively, under the evidence of integration is obtained and is expressed as:
9. the method for diagnosing a malfunction of an anti-recoil mechanism of a gun based on RMSD-DS according to claim 8, wherein said seventh step comprises the steps of:
step 7.1: definition m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted F '(F' ═ X, Y, Z) and F ″ (F ═ X, Y, Z), respectively, giving evidence m 1 And m 2 The DS fusion rule of (1);
in artillery failure mode to give DS fusion rules convenientlyFrame Θ is { X, Y, Z }, m 1 And m 2 For two sets of evidence, the corresponding failure modes are denoted as F '(F' ═ X, Y, Z) and F ″ (F ″ ═ X, Y, Z), respectively, and evidence m is evidence 1 And m 2 The DS fusion rule of (1) is shown as formula (61):
step 7.2: fusing the integration evidence for 3 times according to the DS fusion rule given in the step 7.1, and fusing the integration evidence by using the DS fusion rule to obtain the occurrence probability of the fault mode of the corresponding artillery recoil prevention device;
step 7.3: and (3) obtaining the occurrence probability of the fault modes of all the gun recoil resisting devices according to the traversal in the step (7.2), determining that the fault mode corresponding to the maximum basic probability distribution value is the finally diagnosed fault mode, namely realizing the high-precision and efficient diagnosis of the fault of the gun recoil resisting devices based on an RMSD-DS method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210649020.6A CN115048959B (en) | 2022-06-09 | 2022-06-09 | Method for diagnosing faults of gun anti-squat device based on RMSD-DS |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210649020.6A CN115048959B (en) | 2022-06-09 | 2022-06-09 | Method for diagnosing faults of gun anti-squat device based on RMSD-DS |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115048959A true CN115048959A (en) | 2022-09-13 |
| CN115048959B CN115048959B (en) | 2024-06-21 |
Family
ID=83162215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210649020.6A Active CN115048959B (en) | 2022-06-09 | 2022-06-09 | Method for diagnosing faults of gun anti-squat device based on RMSD-DS |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115048959B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108520266A (en) * | 2018-03-01 | 2018-09-11 | 西北工业大学 | A kind of Time Domain Fusion method for diagnosing faults based on DS evidence theories |
| CN109540520A (en) * | 2018-11-29 | 2019-03-29 | 中国船舶重工集团海装风电股份有限公司 | A kind of rolling bearing fault fusion diagnosis method based on improvement D-S evidence theory |
| US20210003640A1 (en) * | 2019-07-01 | 2021-01-07 | Wuhan University | Fault locating method and system based on multi-layer evaluation model |
| CN113063314A (en) * | 2021-03-23 | 2021-07-02 | 哈尔滨工程大学 | Fault diagnosis method for gun launching system based on SVM (support vector machine) and GA-SVM (genetic algorithm-support vector machine) |
| CN114444585A (en) * | 2022-01-13 | 2022-05-06 | 北京理工大学 | Multi-source information fusion method for conflict evidence |
-
2022
- 2022-06-09 CN CN202210649020.6A patent/CN115048959B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108520266A (en) * | 2018-03-01 | 2018-09-11 | 西北工业大学 | A kind of Time Domain Fusion method for diagnosing faults based on DS evidence theories |
| CN109540520A (en) * | 2018-11-29 | 2019-03-29 | 中国船舶重工集团海装风电股份有限公司 | A kind of rolling bearing fault fusion diagnosis method based on improvement D-S evidence theory |
| US20210003640A1 (en) * | 2019-07-01 | 2021-01-07 | Wuhan University | Fault locating method and system based on multi-layer evaluation model |
| CN113063314A (en) * | 2021-03-23 | 2021-07-02 | 哈尔滨工程大学 | Fault diagnosis method for gun launching system based on SVM (support vector machine) and GA-SVM (genetic algorithm-support vector machine) |
| CN114444585A (en) * | 2022-01-13 | 2022-05-06 | 北京理工大学 | Multi-source information fusion method for conflict evidence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115048959B (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | A novel method for intelligent fault diagnosis of bearing based on capsule neural network | |
| CN111259927B (en) | Rocket engine fault diagnosis method based on neural network and evidence theory | |
| CN109063242B (en) | Guidance tool error identification method based on particle swarm optimization | |
| CN105224872A (en) | A kind of user's anomaly detection method based on neural network clustering | |
| CN112966667B (en) | Method for identifying one-dimensional distance image noise reduction convolution neural network of sea surface target | |
| CN110633790B (en) | Method and system for measuring residual oil quantity of airplane oil tank based on convolutional neural network | |
| CN110082738B (en) | Radar target identification method based on Gaussian mixture and tensor recurrent neural network | |
| CN111126134A (en) | Radar radiation source deep learning identification method based on non-fingerprint signal eliminator | |
| CN108959182B (en) | Small celestial body gravitational field modeling method based on Gaussian process regression | |
| CN109696906B (en) | Underwater robot propeller fault diagnosis method based on wavelet correction Bayes convolution energy | |
| CN110716792B (en) | Target detector and construction method and application thereof | |
| CN112541274A (en) | Missile system efficiency evaluation method based on PCM-TOPSIS method | |
| CN113065223B (en) | Multi-level probability correction method for digital twin model of tower mast cluster | |
| CN106198020A (en) | Wind turbines bearing failure diagnosis method based on subspace and fuzzy C-means clustering | |
| CN113724326B (en) | Monocular vision pose resolving method for taper sleeve target under autonomous aerial refueling scene | |
| CN112132102A (en) | Intelligent fault diagnosis method combining deep neural network with artificial bee colony optimization | |
| CN110701087A (en) | Axial flow compressor pneumatic instability detection method based on single-classification overrun learning machine | |
| CN110223342B (en) | Space target size estimation method based on deep neural network | |
| CN116008671A (en) | Lightning positioning method based on time difference and clustering | |
| CN109669849B (en) | Complex system health state assessment method based on uncertain depth theory | |
| CN115048959A (en) | RMSD-DS-based gun recoil prevention device fault diagnosis method | |
| CN115620172B (en) | Intelligent comprehensive identification method for marine ship target based on cross-domain multi-feature | |
| CN116186586A (en) | Rolling bearing fault diagnosis method based on improved empirical mode decomposition algorithm and optimized deep confidence network | |
| CN110750876A (en) | Bearing data model training and using method | |
| CN115293639A (en) | Battlefield situation studying and judging method based on hidden Markov model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |