CN109683591A - Underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy - Google Patents
Underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy Download PDFInfo
- Publication number
- CN109683591A CN109683591A CN201811609960.2A CN201811609960A CN109683591A CN 109683591 A CN109683591 A CN 109683591A CN 201811609960 A CN201811609960 A CN 201811609960A CN 109683591 A CN109683591 A CN 109683591A
- Authority
- CN
- China
- Prior art keywords
- fault
- signal
- time
- propeller
- hyper
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
- G05B23/0254—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Complex Calculations (AREA)
Abstract
The present invention discloses a kind of underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy, the fault message of two single aspects such as underwater robot speed signal fault message, propeller control signal fault information is organically blended, and then it obtains more fully merging fault message, and the multiple domains fault signatures such as time domain energy, time-frequency entropy are extracted from fusion fault message, for constructing fault sample, finally classified based on support vector domain description algorithm to fault sample, obtains underwater propeller fault degree.The mapping relations of fault signature and fault degree that the invention patent is extracted are unique, and can be realized the classification of propeller fault degree, and nicety of grading reaches 95% or more.
Description
Technical field
The invention belongs to underwater robot propeller technologies, and in particular to one kind is based on fusion signal time domain energy and time-frequency
The underwater propeller fault degree discrimination method of entropy.
Background technique
Propeller is the important component of underwater robot dynamical system.Propeller failure will cause underwater robot fast
The change of the Dynamic Signals such as signal and propeller control voltage signal is spent, and then generates singular behavior in Dynamic Signal.According to
This phenomenon, this patent extract fault signature from dynamic signal singularity behavior, construct fault sample, and build based on fault sample
Vertical failure modes model, diagnoses underwater robot fault degree.Known wavelet energy discrimination method, application No. is
201410705681.1 Chinese patent is single from underwater robot speed signal and propeller control voltage signal two respectively
Aspect extracts Gray-level co-occurrence, recognizes for propeller fault degree, two kinds of fault messages do not organically blend.This
Outside, this kind of method be from signal time domain and frequency domain extraction fault signature, and the mapping relations of gained fault signature and fault degree are not only
One.
Summary of the invention
Goal of the invention: it is an object of the invention to solve the deficiencies in the prior art, provides a kind of based on fusion letter
The underwater propeller fault degree discrimination method of number time domain energy and time-frequency entropy, by underwater robot speed signal fault message,
The fault message of two single aspects such as propeller control signal fault information is merged, and time domain is extracted from fusion feature
Energy feature and time-frequency entropy feature, for constructing fault sample, finally based on support vector domain description algorithm to fault sample into
Row classification, obtains underwater propeller fault degree.
A kind of technical solution: underwater propeller fault degree based on fusion signal time domain energy and time-frequency entropy of the invention
Discrimination method, comprising the following steps:
The first step uses length for L1The time window of (such as can be with value 300) is respectively to underwater robot speed signal
It is intercepted with propeller control voltage change ratio signal;
Second step carries out conventional wavelet decomposition to first step the data obtained, extracts small echo approximation component, to small echo approximation point
Amount carries out conventional amendment Bayes's operation, obtains operation result dSA(n), n is signal data serial number, n=1,2 ..., L1;
Third step is broken down with propeller in n-th of time beat as burnt member Bn, establish failure evidence identification framework Θ
={ B1,B2,…,Bn, the belief assignment function m (B of failure evidence is calculated by formula (1)n), then by m (Bn) be instantiated as
The belief assignment function m of speed signal failure evidenceU(Bn), control signal fault Certainty Factor partition function mC(Bn), it will
mU(Bn)、mC(Bn) bring formula (2)~(3) into and merged, obtain fusion signal fault Certainty Factor partition function mF(Bn);
In formula, dSA(n) Bayes's calculated result, i are corrected for underwater robot Dynamic Signal small echo1For temporary variable, i1
=1,2 ..., N5, N5For time window length, i.e. N5=L1=300, i2And j2For burnt first serial number, i.e. value is 1~L1It is just whole
Number, K5For pilot process variable;
4th step extracts fusion signal time domain energy fault signature FTP:
To fusion signal fault Certainty Factor partition function mF(Bn) carry out convolutional calculation, i.e. mconv(n)=mF(Bn)*mF
(Bn), determine convolutional calculation result mconv(n) all minimum points in carry out the data between two neighboring minimum point
Summation, time domain energy of the acquired results as region between two minimum points, by mconv(n) the time domain energy maximum value in is made
To merge signal time domain energy fault eigenvalue FTP;
5th step extracts fusion signal smoothing puppet Eugene Wigner-Willie and is distributed time-frequency entropy fault signature:
Using smooth and pseudo Wigner-Ville distribution algorithm, as shown in formula (4), fusion signal belief assignment letter is calculated
Number mF(Bn) Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m), calculate the perfume (or spice) of Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m)
Agriculture entropy FTFH, as shown in formula (5)~formula (6), acquired results are as fusion signal time-frequency entropy fault eigenvalue FTFH;
P (n, m)=| SPWVD (n, m) |/∑ ∑ | SPWVD (n, m) | (5)
FTFH=-∑ ∑ p (n, m) log2 p(n,m) (6)
In formula, SPWVD (n, m) is Smoothing Pseudo Eugene Wigner-Willie spectrum, and n is time beat, and n is 1~L1Between integer, m
For band number, m is 1~N4Between integer, N4It often takes 256 and 512 to wait numerical value, such as takes N4=512;
h(k1)、g(l2) it is Gaussian function;
Wherein, k1∈[-(K1-1),(K1- 1)], l2∈[-(L2-1),(L2- 1)], K1、L2Respectively it is not more than (N4)/4、
(N4The maximum integer of)/5, z (n) are the analytic value of speed signal, and z* (n) and z (n) are conjugated, and in this formula, i is imaginary number, i.e. i2
=-1, FTFHTo merge signal time-frequency entropy fault eigenvalue;
6th step will merge signal time domain energy fault signature FTP, fusion signal time-frequency entropy fault signature FTFH, composition event
Hinder sample x=[FTP FTFH]T, it is L by length in step 11Time window function slide to the right, one time beat of every sliding,
One group of underwater robot dynamic signal data is obtained, the first step is repeated to the 6th step, obtains a fault sample, mobile N6When a
Between beat, obtain N6A fault sample, N6For any positive integer, it is the bigger the better in principle for fault sample quantity, such as takes
N6=100;
7th step establishes hyper-sphere model S using conventional support vector domain description method, obtains the hypersphere of hyper-sphere model S
The centre of sphere is C, hyper-sphere model radius is R;
8th step, the classification of propeller fault degree: acquisition propeller fault degree is λqWhen underwater robot dynamic believe
Number, q are propeller fault degree grade, and q=1,2,3 ..., Q, Q is propeller fault degree grade quantity;
According to the first step to the 7th step content, multiple hyper-sphere models are establishedqS, q=1,2,3 ..., Q, and each hypersphere
ModelqS, with the hypersphere centre of sphereqC and radius of hypersphereqR is described;
Test sample is calculated to hyper-sphere modelqThe S centre of sphereqThe generalized distance of CqD passes through formulaqε=qD/qR calculates test specimens
This arrives hyper-sphere modelqThe relative distance of Sqε;Relative distanceqThe corresponding hyper-sphere model of ε minimum valueqFault degree λ representated by Sq,
As propeller fault degree λq。
The utility model has the advantages that the present invention organically blends the fault message of different aspect, and then obtain more fully merging failure
Information, and the multiple domains fault signature such as extraction time domain energy, time-frequency entropy from fusion fault message, fault signature and fault degree
Mapping relations are unique, and the present invention can be realized the classification of propeller fault degree, and nicety of grading reaches 95% or more.
Detailed description of the invention
Fig. 1 is overall flow figure of the invention;
Fig. 2 is underwater robot Dynamic Signal time domain waveform in embodiment;
Fig. 3 is that signal fault characteristic profile is merged in embodiment;
Fig. 4 is propeller fault sample distribution map in embodiment;
Fig. 5 is the fault sample of fault degree 0% in embodiment to the relative distance of each list class hyper-sphere model;
Fig. 6 is the fault sample of fault degree 10% in embodiment to the relative distance of each list class hyper-sphere model;
Fig. 7 is the fault sample of fault degree 20% in embodiment to the relative distance of each list class hyper-sphere model;
Fig. 8 is the fault sample of fault degree 30% in embodiment to the relative distance of each list class hyper-sphere model;
Fig. 9 is the fault sample of fault degree 40% in embodiment to the relative distance of each list class hyper-sphere model.
Specific embodiment
Technical solution of the present invention is described in detail below, but protection scope of the present invention is not limited to the implementation
Example.
As shown in Figure 1,
A kind of underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy of the invention,
The following steps are included:
The first step uses length for L1=300 time window is respectively to underwater robot speed signal and propeller control
Voltage change ratio signal is intercepted;
Second step carries out conventional wavelet decomposition to first step the data obtained, extracts small echo approximation component, to small echo approximation point
Amount carries out conventional amendment Bayes's operation, obtained operation result dsA(n);N is signal data serial number, n=1,2 ..., L1;
Third step is broken down with propeller in n-th of time beat as burnt member Bn, establish failure evidence identification framework Θ
={ B1,B2,…,Bn, the belief assignment function m (B of failure evidence is calculated by formula (1)n), then by m (Bn) be instantiated as
The belief assignment function m of speed signal failure evidenceU(Bn), control signal fault Certainty Factor partition function mC(Bn), it will
mU(Bn)、mC(Bn) bring formula (1)~(3) into and merged, obtain fusion signal fault Certainty Factor partition function mF(Bn);
In formula, dSA(n) Bayes's calculated result, N are corrected for underwater robot Dynamic Signal small echo5For time window length,
That is N5=300;
4th step extracts fusion signal time domain energy fault signature FTP:
To fusion signal fault Certainty Factor partition function mF(Bn) carry out convolutional calculation, i.e. mconv(n)=mF(Bn)*mF
(Bn), determine convolutional calculation result mconv(n) all minimum points in carry out the data between two neighboring minimum point
Summation, time domain energy of the acquired results as region between two minimum points, by mconv(n) the time domain energy maximum value in is made
To merge signal time domain energy fault eigenvalue FTE;
5th step extracts fusion signal smoothing puppet Eugene Wigner-Willie and is distributed time-frequency entropy fault signature:
Using smooth and pseudo Wigner-Ville distribution algorithm, as shown in formula (4), fusion signal belief assignment letter is calculated
Number mF(Bn) Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m), calculate the perfume (or spice) of Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m)
Agriculture entropy FTFH, as shown in formula (5)~formula (6), acquired results are as fusion signal time-frequency entropy fault eigenvalue FTFH;
P (n, m)=| SPWVD (n, m) |/∑ ∑ | SPWVD (n, m) | (5)
FTFH=-∑ ∑ p (n, m) log2 p(n,m) (6)
In formula, SPWVD (n, m) is Smoothing Pseudo Eugene Wigner-Willie spectrum, and n is time beat, and n is 1~L1Between integer, m
For band number, m is 1~N4Between integer, N4Often take 256 and 512 equal numerical value, in the present embodiment, N4=512, h (k1)、g
(l2) it is Gaussian function;
Wherein, k1∈[-(K1-1),(K1- 1)], l2∈[-(L2-1),(L2- 1)], K1、L2Respectively it is not more than (N4)/4、
(N4The maximum integer of)/5, z (n) are the analytic value of speed signal, and z* (n) and z (n) are conjugated, FTFHFor fusion signal time-frequency entropy event
Hinder characteristic value;
6th step will merge signal time domain energy fault signature FTE, fusion signal time-frequency entropy fault signature FTFH, composition event
Hinder sample x=[FTE FTFH]T, time window function is slided to the right, one time beat of every sliding, obtains one group of underwater robot
Dynamic signal data repeats the first step to the 6th step, obtains a fault sample, mobile N6A time beat, until obtaining N6It is a
Fault sample;
7th step establishes hyper-sphere model S using conventional support vector domain description method, obtains the hypersphere of hyper-sphere model S
The centre of sphere is C, hyper-sphere model radius is R;
8th step, the classification of propeller fault degree: acquisition propeller fault degree is λqWhen underwater robot dynamic believe
Number, q are propeller fault degree grade, and q=1,2,3 ..., Q, Q is propeller fault degree grade quantity;
According to the first step to the 7th step content, multiple hyper-sphere models are establishedqS, q=1,2,3 ..., Q, and each hypersphere
ModelqS, with the hypersphere centre of sphereqC and radius of hypersphereqR is described;
Test sample is calculated to hyper-sphere modelqThe S centre of sphereqThe generalized distance of CqD passes through formulaqε=qD/qR calculates test specimens
This arrives hyper-sphere modelqThe relative distance of Sqε;Relative distanceqThe corresponding hyper-sphere model of ε minimum valueqFault degree λ representated by Sq,
As propeller fault degree λq。
Embodiment 1:
As shown in Fig. 2 (a), underwater robot starts since static, underwater robot speed and propeller control voltage
It gradually increases, since the 101st time beat, underwater robot starts to run with the steady state speed of 0.3m/s, at the 250th
At time beat, output loss failure occurs for propeller, and fault degree is respectively 0%, 10%, 20%, 30%, 40%, until
Experiment terminates.As shown in oval frame in Fig. 2 (b) and Fig. 2 (c), underwater robot speed signal, propeller control voltage change ratio
Signal forms singular signal in the 250th~350 time beat.
Use length for L1=300 time window intercepts the speed that the 101st bat is clapped to the 400th in Fig. 2 (b), Fig. 2 (c)
Signal and control voltage change ratio signal data, extract fusion signal time domain energy fault signature, time-frequency from the data of interception
Entropy fault signature constructs fault sample, and time window is moved right 100 time beats, and one time beat of every movement obtains
One group of fault sample amounts to and obtains 500 fault samples, and fault signature distribution is as shown in figure 3, fault sample is distributed as indicated at 4.
In Fig. 3 (a), the corresponding time-frequency entropy feature of larger fault degree is consistently less than time-frequency entropy corresponding compared with glitch degree
Feature, in Fig. 3 (b), it is special that the corresponding time domain energy feature of larger fault degree is consistently greater than the corresponding time domain energy of fault degree
Sign, statistics indicate that, fault degree and fault signature are in dull corresponding relationship in Fig. 3, and a kind of fault signature only corresponds to a kind of failure
The mapping relations of degree, fault signature and fault degree are unique.
From fault sample shown in Fig. 4,50% fault sample is randomly selected as training sample, the failure of residue 50%
Sample is as test sample, shown in fault sample division result table 1.
1 fault sample division result of table
Single class hyper-sphere model that each fault degree is established using the training sample in table 1, by a certain fault degree
Target sample of the corresponding test sample as this kind of fault degree, and using the corresponding test sample of other fault degrees as this
The non-targeted samples of kind fault degree calculate the classification performance of the corresponding single class hyper-sphere model of this kind of fault degree, with index AUC
It is described, the results are shown in Table 2.In table 2, AUC is the area surrounded under ROC curve with reference axis, and ROC is recipient's operation
Indicatrix, AUC more macrotaxonomy device effect is better, and the extreme value of AUC is 1.
The AUC of the fusion signal list class hyper-sphere model of table 2
From Table 2, it can be seen that the corresponding list class hyper-sphere model AUC of fault degree 10% is 0.98, remaining single class hypersphere
Model AUC is 1, is illustrated, the AUC of all single class hyper-sphere models of the present invention is above 0.95, single class hypersphere that the present invention establishes
The classifying quality of model is preferable.
The corresponding test sample of fault degree 0% is calculated to the opposite of the corresponding single class hyper-sphere model of each fault degree
Distance, as a result as shown in Figure 5.In the same way, the corresponding test specimens of fault degree 10%, 20%, 30%, 40% are calculated
This arrives the relative distance of the corresponding single class hyper-sphere model of each fault degree, as a result as shown in Fig. 6~9.In Fig. 5~Fig. 9,
Using the smallest single class hyper-sphere model of relative distance as the affiliated type of the test sample.With correct classification samples number divided by test
Total sample number, result are the nicety of grading of propeller fault degree disaggregated model.As a result, being calculated according to Fig. 5~Fig. 9
Propeller fault degree nicety of grading of the invention is 95.2%, is greater than 95%, classifying quality is preferable.
In addition, extracting time domain energy from speed signal according to the extraction of above-mentioned fault signature and fault degree assorting process
With time-frequency entropy feature, fault sample is constructed, establishes failure modes model, fault degree classification is carried out, obtains single class hyper-sphere model
AUC, as shown in table 3, obtaining fault degree nicety of grading is 90.0%.Using same process, from control voltage change ratio letter
Time domain energy and time-frequency entropy feature are extracted in number, constructs fault sample, establishes failure modes model, carry out fault degree classification,
Single class hyper-sphere model AUC is obtained, as shown in table 3, obtaining fault degree nicety of grading is 72.4%.
The AUC of 3 speed signal of table and control signal list class hyper-sphere model
Contrast table 2 and 3 data of table are it is found that fusion signal list class hyper-sphere model AUC and speed signal list class hyper-sphere model AUC
It compares, fault degree 10%, 20% corresponding AUC are larger, remaining is equal, and fusion signal list class hyper-sphere model AUC and control are believed
Number list class hyper-sphere model AUC is compared, and the corresponding AUC of fault degree 40% is equal, remaining is all larger, illustrates to merge signal list class super
The classification performance of spherical model is better than speed signal list class hyper-sphere model and control signal list class hyper-sphere model.
In addition, merging the nicety of grading 95.2% of signal fault degree classification model in terms of nicety of grading, it is greater than speed
The nicety of grading 90.0% of signal fault degree classification model, greater than the nicety of grading of control signal fault degree classification model
72.4%.
Claims (1)
1. a kind of underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy, feature exist
In: the following steps are included:
The first step uses length for L1Time window respectively to underwater robot speed signal and propeller control voltage change ratio
Signal is intercepted;
Second step carries out conventional wavelet decomposition to first step the data obtained, extracts small echo approximation component sA(n), to small echo approximation point
Measure sA(n) conventional amendment Bayes's operation is carried out, operation result d is obtainedsA(n);
Third step is broken down with propeller in n-th of time beat as burnt member Bn, establish failure evidence identification framework Θ=
{B1,B2,L,Bn, the belief assignment function m (B of failure evidence is calculated by formula (1)n), then by m (Bn) it is instantiated as speed
The belief assignment function m of signal fault evidenceU(Bn), control signal fault Certainty Factor partition function mC(Bn), by mU
(Bn)、mC(Bn) bring formula (2)~(3) into and merged, obtain fusion signal fault Certainty Factor partition function mF(Bn);
In formula, dSA(n) Bayes's calculated result, i are corrected for underwater robot Dynamic Signal small echo1For temporary variable, i1=1,
2 ..., N5, N5For time window length, that is, N5=L1, i2And j2For burnt first serial number, i.e. value is 1~L1Positive integer, K5For centre
Process variable;
4th step extracts fusion signal time domain energy fault signature FTP:
To fusion signal fault Certainty Factor partition function mF(Bn) carry out convolutional calculation, i.e. mconv(n)=mF(Bn)*mF(Bn),
Determine convolutional calculation result mconv(n) all minimum points in, the data between two neighboring minimum point are summed,
Time domain energy of the acquired results as region between two minimum points, by mconv(n) the time domain energy maximum value in, which is used as, melts
Close signal time domain energy fault eigenvalue FTP;
5th step extracts fusion signal smoothing puppet Eugene Wigner-Willie and is distributed time-frequency entropy fault signature:
Using smooth and pseudo Wigner-Ville distribution algorithm, as shown in formula (4), fusion signal belief assignment function m is calculatedF
(Bn) Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m), calculate the Shannon entropy of Smoothing Pseudo Eugene Wigner-Willie spectrum SPWVD (n, m)
FTFH, as shown in formula (5)~formula (6), acquired results are as fusion signal time-frequency entropy fault eigenvalue FTFH;
P (n, m)=| SPWVD (n, m) |/∑ ∑ | SPWVD (n, m) | (5)
FTFH=-∑ ∑ p (n, m) log2p(n,m) (6)
In formula, SPWVD (n, m) is Smoothing Pseudo Eugene Wigner-Willie spectrum, and n is time beat, and n is 1~L1Between integer, m be frequency
Band serial number, m are 1~N4Between integer, h (k1)、g(l2) it is Gauss function;
Wherein, k1∈[-(K1-1),(K1- 1)], l2∈[-(L2-1),(L2- 1)], K1、L2Respectively it is not more than (N4)/4、(N4)/5
Maximum integer, z (n) is the analytic value of speed signal, and z* (n) and z (n) are conjugated, and in this formula, i is imaginary number, i.e. i2=-1,
FTFHTo merge signal time-frequency entropy fault eigenvalue;
6th step will merge signal time domain energy fault signature FTP, fusion signal time-frequency entropy fault signature FTFH, form failure sample
This x=[FTP FTFH]T, it is L by length in step 11Time window function slide to the right, one time beat of every sliding, obtain
One group of underwater robot dynamic signal data repeats the first step to the 6th step, obtains a fault sample, mobile N6Segmentum intercalaris when a
It claps, obtains N6A fault sample, N6For any positive integer;
7th step establishes hyper-sphere model S using conventional support vector domain description method, obtains the hypersphere centre of sphere of hyper-sphere model S
It is R for C, hyper-sphere model radius;
8th step, the classification of propeller fault degree: acquisition propeller fault degree is λqWhen underwater robot Dynamic Signal number
According to q is propeller fault degree grade, and q=1,2,3 ..., Q, Q is propeller fault degree grade quantity;
According to the first step to the 7th step content, multiple hyper-sphere models are establishedqS, q=1,2,3 ..., Q, and each hyper-sphere modelqS, with the hypersphere centre of sphereqC and radius of hypersphereqR is described;
Test sample is calculated to hyper-sphere modelqThe S centre of sphereqThe generalized distance of CqD passes through formulaqε=qD/qR calculates test sample and arrives
Hyper-sphere modelqThe relative distance of Sqε;Relative distanceqThe corresponding hyper-sphere model of ε minimum valueqFault degree λ representated by Sq, as
Propeller fault degree λq。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811609960.2A CN109683591B (en) | 2018-12-27 | 2018-12-27 | Underwater propeller fault degree identification method based on fusion signal time domain energy and time-frequency entropy |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811609960.2A CN109683591B (en) | 2018-12-27 | 2018-12-27 | Underwater propeller fault degree identification method based on fusion signal time domain energy and time-frequency entropy |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109683591A true CN109683591A (en) | 2019-04-26 |
CN109683591B CN109683591B (en) | 2021-03-19 |
Family
ID=66190017
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811609960.2A Active CN109683591B (en) | 2018-12-27 | 2018-12-27 | Underwater propeller fault degree identification method based on fusion signal time domain energy and time-frequency entropy |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109683591B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110716532A (en) * | 2019-10-17 | 2020-01-21 | 哈尔滨工程大学 | Underwater robot propeller weak fault identification method based on wavelet packet energy and FFT |
CN112015160A (en) * | 2019-05-31 | 2020-12-01 | 北京新能源汽车股份有限公司 | Fault temperature determination method and device |
CN112034826A (en) * | 2020-09-11 | 2020-12-04 | 江苏科技大学 | Method for identifying fault degree of underwater propeller based on least square method |
CN112051836A (en) * | 2020-09-11 | 2020-12-08 | 江苏科技大学 | Underwater robot propeller state monitoring method based on multi-core model |
CN113655778A (en) * | 2021-08-16 | 2021-11-16 | 江苏科技大学 | Underwater propeller fault diagnosis system and method based on time-frequency energy |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104503432A (en) * | 2014-11-27 | 2015-04-08 | 哈尔滨工程大学 | Autonomous underwater robot fault identification method based on wavelet energy |
CN107133376A (en) * | 2017-03-27 | 2017-09-05 | 哈尔滨工程大学 | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model |
CN107132760A (en) * | 2017-03-27 | 2017-09-05 | 哈尔滨工程大学 | The underwater robot state monitoring method described based on fuzzy support vector domain |
CN109033965A (en) * | 2018-06-22 | 2018-12-18 | 江苏科技大学 | A kind of underwater robot propeller failure time-frequency characteristics Enhancement Method |
-
2018
- 2018-12-27 CN CN201811609960.2A patent/CN109683591B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104503432A (en) * | 2014-11-27 | 2015-04-08 | 哈尔滨工程大学 | Autonomous underwater robot fault identification method based on wavelet energy |
CN107133376A (en) * | 2017-03-27 | 2017-09-05 | 哈尔滨工程大学 | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model |
CN107132760A (en) * | 2017-03-27 | 2017-09-05 | 哈尔滨工程大学 | The underwater robot state monitoring method described based on fuzzy support vector domain |
CN109033965A (en) * | 2018-06-22 | 2018-12-18 | 江苏科技大学 | A kind of underwater robot propeller failure time-frequency characteristics Enhancement Method |
Non-Patent Citations (4)
Title |
---|
张铭钧,刘维新,殷宝吉,王玉甲: "自主式水下机器人故障特征增强方法", 《哈尔滨工程大学学报》 * |
张铭钧,殷宝吉,刘维新,王玉甲: "随机干扰下AUV推进器故障特征提取与融合", 《华中科技大学学报》 * |
殷宝吉: "水下机器人推进器故障特征分离与故障程度辨识方法研究", 《中国博士学位论文全文数据库信息科技辑》 * |
沈长青: "旋转机械设备关键部件故障诊断与预测方法研究", 《中国博士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112015160A (en) * | 2019-05-31 | 2020-12-01 | 北京新能源汽车股份有限公司 | Fault temperature determination method and device |
CN112015160B (en) * | 2019-05-31 | 2021-10-22 | 北京新能源汽车股份有限公司 | Fault temperature determination method and device |
CN110716532A (en) * | 2019-10-17 | 2020-01-21 | 哈尔滨工程大学 | Underwater robot propeller weak fault identification method based on wavelet packet energy and FFT |
CN112034826A (en) * | 2020-09-11 | 2020-12-04 | 江苏科技大学 | Method for identifying fault degree of underwater propeller based on least square method |
CN112051836A (en) * | 2020-09-11 | 2020-12-08 | 江苏科技大学 | Underwater robot propeller state monitoring method based on multi-core model |
CN113655778A (en) * | 2021-08-16 | 2021-11-16 | 江苏科技大学 | Underwater propeller fault diagnosis system and method based on time-frequency energy |
CN113655778B (en) * | 2021-08-16 | 2024-03-26 | 江苏科技大学 | Underwater propeller fault diagnosis system and method based on time-frequency energy |
Also Published As
Publication number | Publication date |
---|---|
CN109683591B (en) | 2021-03-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109683591A (en) | Underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy | |
CN104794368A (en) | Rolling bearing fault classifying method based on FOA-MKSVM (fruit fly optimization algorithm-multiple kernel support vector machine) | |
CN107462785A (en) | The more disturbing signal classifying identification methods of the quality of power supply based on GA SVM | |
CN109816032A (en) | Zero sample classification method and apparatus of unbiased mapping based on production confrontation network | |
CN105403873B (en) | Target's feature-extraction method based on Fourier Transform of Fractional Order | |
CN116029219B (en) | Aerodynamic heat prediction method, device, equipment and storage medium for aircraft | |
Liu et al. | GanDef: A GAN based adversarial training defense for neural network classifier | |
CN109696906A (en) | Underwater robot propeller method for diagnosing faults based on small echo amendment Bayes's convolution energy | |
CN116361718A (en) | Single-phase earth fault line selection method based on improved variational modal decomposition and dispersion entropy | |
Zhang | Use density-based spatial clustering of applications with noise (DBSCAN) algorithm to identify galaxy cluster members | |
CN109633270A (en) | The identification of fault energy zone boundary and feature extracting method based on instantaneous spectrum entropy and noise energy difference | |
CN105913199A (en) | Energy, economy and environment coordination degree calculation method and device based on ellipsoidal model | |
CN103793747A (en) | Sensitive information template construction method in network content safety management | |
McClintick et al. | Countering physical eavesdropper evasion with adversarial training | |
Liu et al. | Community detection enhancement using non-negative matrix factorization with graph regularization | |
Zhang et al. | Adversarial Domain Generalization Defense for Automatic Modulation Classification | |
Mengmeng et al. | Signal sorting using teaching-learning-based optimization and random forest | |
Feng et al. | FCGCN: Feature Correlation Graph Convolution Network for Few-Shot Individual Identification | |
CN106911512A (en) | Link Forecasting Methodology and system based on game in commutative figure | |
Liu et al. | Radar signal waveform recognition based on convolutional denoising autoencoder | |
Xu et al. | Classification method of marine tourism resource of least square support vector machines based on particle swarm algorithm | |
Dou et al. | Feature GANs: A Model for Data Enhancement and Sample Balance of Foreign Object Detection in High Voltage Transmission Lines | |
CN110110779A (en) | Virtual sample generation method based on Density Estimator and Copula function | |
Su et al. | A Generalized Fuzzy T-norm Formulation of Fuzzy Modularity for Community Detection in Social Networks. | |
CN108491563A (en) | A kind of signal envelope extracting method based on sparse reconstruct optimization algorithm |
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 |