CN105866664A - Intelligent fault diagnosis method for analog circuit based on amplitude frequency features - Google Patents
Intelligent fault diagnosis method for analog circuit based on amplitude frequency features Download PDFInfo
- Publication number
- CN105866664A CN105866664A CN201610447694.2A CN201610447694A CN105866664A CN 105866664 A CN105866664 A CN 105866664A CN 201610447694 A CN201610447694 A CN 201610447694A CN 105866664 A CN105866664 A CN 105866664A
- Authority
- CN
- China
- Prior art keywords
- amplitude
- analog circuit
- fault
- parameter
- value
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/316—Testing of analog circuits
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/285—Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Abstract
The invention provides an intelligent fault diagnosis method for an analog circuit based on amplitude frequency features. The intelligent fault diagnosis method for the analog circuit based on the amplitude frequency features comprises the following steps of: a) applying a stimulus signal to a to-be-diagnosed circuit, collecting the amplitude frequency response information outputted by the to-be-diagnosed circuit and taking as fault feature information; b) analyzing and treating the amplitude frequency response information and acquiring an optimized sample input parameter of a classifier according to the amplitude frequency response information; and c) inputting the optimized sample input parameter into the corresponding classifier, for classifying and identifying the fault, thereby acquiring a final diagnosis result. The intelligent fault diagnosis method provided by the invention has the beneficial effects that the fault of the analog circuit can be automatically diagnosed according to the intelligent fault diagnosis method for the analog circuit based on the amplitude frequency features and the accuracy and effect of the fault diagnosis for the analog circuit are promoted.
Description
Technical field
The invention belongs to ic test technique field, analog circuit can be improved more particularly to a kind of
Fault diagnosis rate and efficiency, and the analog circuit based on amplitude-frequency characteristic of more fault type can be diagnosed to be
Intelligent fault diagnostic method.
Background technology
Along with the development of microelectric technique, extensive hybrid digital-analog integrated circuit is extensively applied in all trades and professions,
The fault that research is diagnosed to be in analog circuit the most accurately and rapidly becomes in the urgent need to address in Practical Project
Problem；The features such as non-linear, the tolerance of element, the failure mode diversity due to analog circuit cause
Analog Circuit Fault Diagnosis Technology slower development, traditional Troubleshooting Theory and method in Practical Project very
Difficulty gets a desired effect.
Analog circuit is that to process analog signal, i.e. time and amplitude be all the electronic circuit of continuous signal, its
Fault diagnosis shows the feature of many complexity.Although the most many experts and scholars are to analog circuit fault
Diagnosis expands substantial amounts of research, the method emerging many analog circuit fault diagnosings, but just because of mould
The own characteristic intending circuit causes existing analog-circuit fault diagnosis method the most perfect.By to existing
Domestic and international analog circuit fault diagnosing Research foundation on be analyzed and sum up, find existing simulation electricity
Road method for diagnosing faults mainly there is also following problems:
(1) accuracy of analog circuit fault diagnosing need to improve further.Due in analog circuit element
There is tolerance so that the feature of some fault mode is closely similar, thus add the fuzzy of failure modes
Property and uncertainty.The fault mode that existing method is typically these to be difficult to differentiate between merges into a class fault
Treat or by the diagnosable element in Testability Analysis the diagnostic circuit to circuit, therefore these methods
Correct diagnosis to all fault modes can't be fully achieved.
(2) determination of analog circuit fault model is unreasonable, and fault arranges interval excessive, and neither one is unified
Standard, cause the not preciseness of analog circuit fault diagnosing.According to the fault degree of analog circuit, have soft
Fault and hard fault point.It is typically with the mode mould of one resistance of serial or parallel connection currently for hard fault
Intend open circuit or the short circuit of circuit, as former in the short circuit of the two ends analog element of fault element with the resistor coupled in parallel of 1
Barrier；With the resistant series of a 100M at the open fault of the branch road analog element at fault element place.Generally
Thinking that device parameter values offset by the principle of the soft fault that its range of tolerable variance causes is that component parameters is more than or little
Within 10 times of its nominal value.Existing method has the biggest randomness arranging fault mode when,
The soft fault that deviation nominal value is bigger can there be is preferable diagnosis effect, if but soft fault being positioned proximate to nominal
If value, the precision of analog circuit fault diagnosing will be substantially reduced.
(3) in analog circuit, nonlinear circuit is widely present, and adds the reason such as tolerance of element, simulation
The output response of circuit the most dynamically changes so that some fault signature of circuit shows very
Similar feature, and only have small difference in local, it is very difficult to distinguishing, this makes fault diagnosis be difficult to.
Therefore, this link of analog circuit fault feature extraction is particularly important, and it is the fault diagnosis effect obtained
One of crucial.Feature extracting method great majority method based on signal transacting at present, to analog circuit local letter
Number carrying out feature extraction, but effect is barely satisfactory, similar fault is still difficult to differentiate between, the mistaken diagnosis of fuzzy fault
Rate is the highest.
(4) analog circuit fault sample acquisition difficulty, analog circuit fault diagnosing is a typical small sample
Pattern recognition problem, the shortage of fault sample become have a strong impact on Analog Circuit Fault Diagnosis Technology development
One of major reason.Existing Fault Classification is not fine to the effect of small sample problem.
Therefore, it is necessary to a kind of fault diagnosis rate that can improve analog circuit and efficiency are provided, and can
It is diagnosed to be the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic of more fault type.
Summary of the invention
It is an object of the invention to provide a kind of fault diagnosis rate that can improve analog circuit and efficiency, and
The analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic of more fault type can be diagnosed to be.
Technical scheme is as follows: a kind of analog circuit Intelligent fault diagnosis side based on amplitude-frequency characteristic
Method comprises the steps:
A, treat diagnostic circuit and apply pumping signal, and treat that the amplitude-frequency response that diagnostic circuit export is believed described in gathering
Breath is as fault characteristic information；
B, described amplitude-frequency response information it is analyzed and processes, and dividing according to described amplitude-frequency response information acquisition
The optimization sample input parameter of class device；
C, by described optimization sample input parameter input corresponding described grader carry out failure modes identification,
And obtain last diagnostic result.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
In step a, described pumping signal be amplitude be the sine voltage signal of 1V.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Described step b specifically includes following steps:
B1, utilize principal component analytical method (Principal Component Analysis, PCA) to described width
Frequently response message carries out feature extraction and dimension-reduction treatment；
B2, the described amplitude-frequency response information after feature extraction and dimension-reduction treatment is normalized；
B3, utilize particle cluster algorithm (particle swarm optimization, PSO) to normalization after institute
State amplitude-frequency response information to be optimized, thus obtain the optimization sample input parameter of grader.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
In step b2, choose multiple attribute according to contribution rate and constitute fault signature, and described fault signature is carried out
Normalized.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Described normalized is that all of data are converted into the number between [0,1], and, described normalized institute
The data normalization function used is as follows:
Wherein, xmaxFor the maximum number in data sequence；xminFor the minimum number in sequence, xkWith x 'kPoint
Value before and after Wei not normalizing.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Described step b3 specifically includes following steps:
Determine fitness function, training sample is carried out the accuracy rate under cross validation as Particle Swarm Optimization
Fitness function value in method；
Population initializes, the population that stochastic generation scale is moderate in solution space, and population is individual
Represent the parameter of grader；
Calculate particle fitness value, punishment parameter C and the kernel functional parameter σ of grader are set, sample set is defeated
Enter grader to be trained, obtain the discrimination of test sample, according to the evaluation function of grader classification performance
Calculate particle fitness；
Determine individual extreme value PbestWith global extremum gbest, the particle fitness value after updating and described individual pole
Value PbestCorresponding adaptive value compares, if excellent, updates described individual extreme value Pbest, otherwise retain initial value；Will more
Described individual extreme value P of each particle after XinbestWith described global extremum gbestRelatively, if excellent, update described
Global extremum gbest, otherwise retain initial value；
The position of more new particle and velocity information；
Judge whether to meet end condition, the most whether meet other end conditions of maximum iteration time or setting,
If met, output category device parameter, this parameter is optimized parameter, and algorithm terminates；Otherwise return calculating
Particle fitness value step, until meeting end condition, exports optimum classifier parameters value.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Described grader uses SVMs (support vector machine, SVM).
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
The optimization sample input parameter of described grader include punishing accordingly with described SVMs parameter C and
Kernel functional parameter σ.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Multiple graders are constructed according to one-to-many combination in step c.
In the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides,
Before step a, also include step c: treat that diagnostic circuit determines the fault mode of circuit according to described.
The beneficial effects of the present invention is: described analog circuit Intelligent fault diagnosis side based on amplitude-frequency characteristic
In method, collection analog circuit amplitude-frequency response, as fault signature, then utilizes principal component analysis to carry out fault special
Levy extraction and dimension-reduction treatment, such that it is able to the redundancy efficiently reduced in fault signature and interference component.And,
Fault grader uses SVMs, and uses particle swarm optimization algorithm optimizing to obtain described support
The kernel functional parameter of vector machine and punishment parameter, thus improve the precision of analog circuit fault diagnosing.Therefore,
Described analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic can diagnose analog circuit fault automatically,
And significantly improve precision and the effect of analog circuit fault diagnosing.
Accompanying drawing explanation
Fig. 1 is the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides
Schematic flow sheet；
Fig. 2 is the analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic that the embodiment of the present invention provides
FB(flow block)；
Fig. 3 is the electrical block diagram of four high guaily unit biquadratic high-pass filter；
Fig. 4 is step S3 of analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic shown in Fig. 2
FB(flow block)；
Fig. 5 is particle group optimizing SVMs schematic flow sheet in step S3 shown in Fig. 4.
Detailed description of the invention
In order to make the purpose of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and reality
Execute example, the present invention is further elaborated.Only should be appreciated that specific embodiment described herein
Only in order to explain the present invention, it is not intended to limit the present invention.
The description of specific distinct unless the context otherwise, the element in the present invention and assembly, quantity both can be single
Presented in individual, it is also possible to presented in multiple, this is not defined by the present invention.In the present invention
Although step arranged with label, but be not used to limit the precedence of step, unless specifically
Based on understanding that the order of step or the execution of certain step need other steps, otherwise step is the most secondary
Sequence is adjustable in.It is appreciated that term "and/or" used herein relates to and contains to be associated
One or more of any and all possible combination in Listed Items.
It is simulation based on the amplitude-frequency characteristic electricity that the embodiment of the present invention provides please refer to Fig. 1 and Fig. 2, Fig. 1
The schematic flow sheet of road Intelligent fault diagnostic method, Fig. 2 is the special based on amplitude-frequency of embodiment of the present invention offer
The FB(flow block) of the analog circuit Intelligent fault diagnostic method levied.It is based on amplitude-frequency characteristic that the present invention provides
Analog circuit Intelligent fault diagnostic method 100 comprises the steps:
S1, basis treat that diagnostic circuit determines the fault mode of circuit.
Specifically, according to the model of electronic devices and components and the performance treating diagnostic circuit, it may be determined that described follow-up
The contingent fault mode of deenergizing.
Such as, as a example by the four high guaily unit biquadratic high-pass filter shown in Fig. 3, the nominal of each element in circuit
As shown in Figure 3, wherein resistance and electric capacity respectively have the tolerance of 5% to value.Element deviate its nominal value ± 50%
Time, circuit generation soft fault, i.e. value during element generation soft fault should [50%X,
95%X) ∪ (105%X, 150%X] interval in (X is the nominal value of element).Table 1 gives four high guaily unit double two
Secondary high-pass filter normal value, soft fault value and the fault category of correspondence.Wherein,Represent that element there occurs partially
Major break down,Represent that element there occurs fault less than normal, including having 13 kinds of fault modes within unfaulty conditions.
Table 1 fault mode arranges table
S2, treat diagnostic circuit apply pumping signal, and gather described in treat the amplitude-frequency response that diagnostic circuit exports
Information is as fault characteristic information.
Specifically, give and treat that the input of diagnostic circuit applies sine voltage signal that an amplitude is 1V as excitation
Signal, carries out Parameter analysis (AC Sweep), gathers amplitude-frequency respectively at the described output treating diagnostic circuit and rings
Induction signal is as the fault signature of circuit.And, arranging initial frequency is 1kHz, and termination frequency is 25kHz,
All fault modes are carried out the Monte Carlo analysis of 50 times, obtain every kind of event by 60 sampled points respectively
Barrier pattern is respectively provided with 50 samples of 60 attributes.
S3, described amplitude-frequency response information is analyzed and processes, and according to described amplitude-frequency response information acquisition
The optimization sample input parameter of grader.
Specifically, refer to Fig. 4, be analog circuit Intelligent fault based on amplitude-frequency characteristic diagnosis shown in Fig. 2
The FB(flow block) of step S3 of method.Described step S3 comprises the steps:
S31, utilize principal component analytical method that described amplitude-frequency response information is carried out feature extraction and dimension-reduction treatment；
S32, the described amplitude-frequency response information after feature extraction and dimension-reduction treatment is normalized；
S33, utilize particle cluster algorithm that the described amplitude-frequency response information after normalization is optimized, thus obtain
The optimization sample input parameter of grader.
In step S31 and step S32, owing to the sample of described amplitude-frequency response information containing substantial amounts of redundancy
Composition, influences whether final diagnostic result and performance, needs the sample principal component analytical method that will obtain
Carry out feature extraction and dimension-reduction treatment.Preferably, during using principal component analytical method, according to contribution
Rate is chosen multiple attribute and is constituted fault signature, and is normalized described fault signature.Such as, choosing
Front 7 attributes taking contribution rate maximum constitute fault signature, and are normalized fault eigenvalue.
And, described normalized is that all of data are converted into the number between [0,1], its objective is to cancel
Order of magnitude difference between each dimension data, it is to avoid cause identification effect because inputoutput data order of magnitude difference is relatively big
Fruit is the best.
In the present embodiment, the data normalization function that described normalized is used is as follows:
Wherein, xmaxFor the maximum number in data sequence；xminFor the minimum number in sequence, xkWith x 'kPoint
Value before and after Wei not normalizing.
Such as, if as a example by the four high guaily unit biquadratic high-pass filter shown in Fig. 3, permissible according to step S31
Obtain 50 samples with 7 attributes.
In step S33, the fault signature after step S32 being processed is divided into training sample and test sample two
Part, and described training sample input SVMs is trained, prop up with particle cluster algorithm optimizing simultaneously
Hold kernel functional parameter and the punishment parameter of vector machine.
Such as, 50 samples step S31 obtained are divided into two parts: front 30 samples are as training sample
This, rear 20 samples are as test sample, owing to there being 13 kinds of fault modes, finally give 390 training
Sample and 260 test samples.Sample is inputted SVMs be trained, use particle cluster algorithm simultaneously
The kernel functional parameter of Support Vector Machines Optimized and punishment parameter.Using particle cluster algorithm produce initial population as
SVMs parameter input supporting vector machine model be trained and test, by update particle position and
Velocity information iteration optimizing produces parameter and population of future generation, until meeting the termination bar arranged in particle cluster algorithm
Part, finally gives punishment parameter C and the kernel functional parameter σ of optimum.
Specifically, refer to Fig. 5, be that in step S3 shown in Fig. 4, particle group optimizing SVMs flow process is shown
It is intended to.Described step S33 specifically includes following steps:
Determine fitness function, training sample is carried out the accuracy rate under cross validation as Particle Swarm Optimization
Fitness function value in method；
Population initializes, the population that stochastic generation scale is moderate in solution space, and population is individual
Represent the parameter of grader；
Calculate particle fitness value, punishment parameter C and the kernel functional parameter σ of grader are set, sample set is defeated
Enter grader to be trained, obtain the discrimination of test sample, according to the evaluation function of grader classification performance
Calculate particle fitness；
Determine individual extreme value PbestWith global extremum gbest, the particle fitness value after updating and described individual pole
Value PbestCorresponding adaptive value compares, if excellent, updates described individual extreme value Pbest, otherwise retain initial value；Will more
Described individual extreme value P of each particle after XinbestWith described global extremum gbestRelatively, if excellent, update described
Global extremum gbest, otherwise retain initial value；
The position of more new particle and velocity information；
Judge whether to meet end condition, the most whether meet other end conditions of maximum iteration time or setting,
If met, output category device parameter, this parameter is optimized parameter, and algorithm terminates；Otherwise return calculating
Particle fitness value step, until meeting end condition, exports optimum classifier parameters value.
S4, by described optimization sample input parameter input corresponding described grader carry out failure modes identification,
And obtain last diagnostic result.
Specifically, in step s 4, described grader uses SVMs, the most described grader excellent
Change sample input parameter and include punishing accordingly with described SVMs parameter C and kernel functional parameter σ.
The parameter being additionally, since described SVMs is relatively big on the impact of final classification results, only finds
Optimized parameter could improve fault diagnosis precision.Therefore, by using described particle swarm optimization algorithm optimizing to obtain
To punishment parameter C and the kernel functional parameter of described SVMs, analog circuit fault diagnosing can be improve
Precision.
In the present embodiment, analog circuit fault diagnosing is classification problem more than, needs to construct multi-categorizer,
The present invention uses one-to-many combination to construct multiple graders.Such as K is classified, the most first construct
Individual grader, then will result be voted by sample input model, what poll was most is final classification results.
The most such as, as a example by the four high guaily unit biquadratic high-pass filter shown in Fig. 3, as shown in table 2, only 2
4 sample DE in class fault, fault diagnosis rate reaches 98.5%.
Table 2 fault diagnosis result
Compared to prior art, analog circuit Intelligent fault based on the amplitude-frequency characteristic diagnosis that the present invention provides
In method 100, collection analog circuit amplitude-frequency response, as fault signature, then utilizes principal component analysis to carry out
Fault signature extracts and dimension-reduction treatment, such that it is able to the redundancy efficiently reduced in fault signature and interference component.
And, fault grader uses SVMs, and uses particle swarm optimization algorithm optimizing to obtain institute
State kernel functional parameter and the punishment parameter of SVMs, thus improve the precision of analog circuit fault diagnosing.
Therefore, analog circuit Intelligent fault diagnostic method 100 based on amplitude-frequency characteristic can diagnose analog circuit automatically
Fault, and significantly improve precision and the effect of analog circuit fault diagnosing.
It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment,
And without departing from the spirit or essential characteristics of the present invention, it is possible to realize in other specific forms
The present invention.Therefore, no matter from the point of view of which point, embodiment all should be regarded as exemplary, and right and wrong
Restrictive, the scope of the present invention is limited by claims rather than described above, it is intended that will fall
All changes in the implication of equivalency and scope of claim are included in the present invention.Should will not weigh
Any reference during profit requires is considered as limiting involved claim.
Moreover, it will be appreciated that although this specification is been described by according to embodiment, but the most each enforcement
Mode only comprises an independent technical scheme, and this narrating mode of specification is only for clarity sake,
Those skilled in the art should be using specification as an entirety, and the technical scheme in each embodiment can also be through
Appropriately combined, form other embodiments that it will be appreciated by those skilled in the art that.
Claims (10)
1. an analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic, it is characterised in that: include
Following steps:
A, treat diagnostic circuit and apply pumping signal, and treat that the amplitude-frequency response that diagnostic circuit export is believed described in gathering
Breath is as fault characteristic information；
B, described amplitude-frequency response information it is analyzed and processes, and dividing according to described amplitude-frequency response information acquisition
The optimization sample input parameter of class device；
C, by described optimization sample input parameter input corresponding described grader carry out failure modes identification,
And obtain last diagnostic result.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 1,
It is characterized in that: in step a, described pumping signal be amplitude be the sine voltage signal of 1V.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 1,
It is characterized in that: described step b specifically includes following steps:
B1, utilize principal component analytical method that described amplitude-frequency response information is carried out feature extraction and dimension-reduction treatment；
B2, the described amplitude-frequency response information after feature extraction and dimension-reduction treatment is normalized；
B3, utilize particle cluster algorithm that the described amplitude-frequency response information after normalization is optimized, thus obtain
The optimization sample input parameter of grader.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 3,
It is characterized in that: in step b2, choose multiple attribute according to contribution rate and constitute fault signature, and to described
Fault signature is normalized.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 4,
It is characterized in that: described normalized is that all of data are converted into the number between [0,1], and, described
The data normalization function that normalized is used is as follows:
Wherein, xmaxFor the maximum number in data sequence；xminFor the minimum number in sequence, xkWith x 'kRespectively
For the value before and after normalization.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 3,
It is characterized in that: described step b3 specifically includes following steps:
Determine fitness function, training sample is carried out the accuracy rate under cross validation as Particle Swarm Optimization
Fitness function value in method；
Population initializes, the population that stochastic generation scale is moderate in solution space, and population is individual
Represent the parameter of grader；
Calculate particle fitness value, punishment parameter C and the kernel functional parameter σ of grader are set, sample set is defeated
Enter grader to be trained, obtain the discrimination of test sample, according to the evaluation function of grader classification performance
Calculate particle fitness；
Determine individual extreme value PbestWith global extremum gbest, the particle fitness value after updating and described individual pole
Value PbestCorresponding adaptive value compares, if excellent, updates described individual extreme value Pbest, otherwise retain initial value；Will more
Described individual extreme value P of each particle after XinbestWith described global extremum gbestRelatively, if excellent, update described
Global extremum gbest, otherwise retain initial value；
The position of more new particle and velocity information；
Judge whether to meet end condition, the most whether meet other end conditions of maximum iteration time or setting,
If met, output category device parameter, this parameter is optimized parameter, and algorithm terminates；Otherwise return calculating
Particle fitness value step, until meeting end condition, exports optimum classifier parameters value.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 1,
It is characterized in that: described grader uses SVMs.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 7,
It is characterized in that: the optimization sample input parameter of described grader includes with described SVMs accordingly
Punishment parameter C and kernel functional parameter σ.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 7,
It is characterized in that: in step c, construct multiple graders according to one-to-many combination.
Analog circuit Intelligent fault diagnostic method based on amplitude-frequency characteristic the most according to claim 1,
It is characterized in that: before step a, also include step c: treat that diagnostic circuit determines the event of circuit according to described
Barrier pattern.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610447694.2A CN105866664A (en) | 2016-06-20 | 2016-06-20 | Intelligent fault diagnosis method for analog circuit based on amplitude frequency features |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610447694.2A CN105866664A (en) | 2016-06-20 | 2016-06-20 | Intelligent fault diagnosis method for analog circuit based on amplitude frequency features |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105866664A true CN105866664A (en) | 2016-08-17 |
Family
ID=56650155
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610447694.2A Pending CN105866664A (en) | 2016-06-20 | 2016-06-20 | Intelligent fault diagnosis method for analog circuit based on amplitude frequency features |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105866664A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778694A (en) * | 2017-01-18 | 2017-05-31 | 北京工业大学 | A kind of gear transmission noises analysis method based on set empirical mode decomposition and SVMs |
| CN107202950A (en) * | 2017-05-12 | 2017-09-26 | 南京邮电大学 | Analog-circuit fault diagnosis method based on LMD approximate entropies and SVM |
| CN110096927A (en) * | 2018-01-30 | 2019-08-06 | 西安交通大学 | Contactor diagnostic method and diagnostic system based on particle group optimizing support vector machines |
| CN111461208A (en) * | 2020-03-31 | 2020-07-28 | 贵州电网有限责任公司 | Development scale prediction method and system suitable for distributed energy supply system |
| CN113487019A (en) * | 2021-07-06 | 2021-10-08 | 湖南第一师范学院 | Circuit fault diagnosis method and device, computer equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH02254377A (en) * | 1989-03-29 | 1990-10-15 | Sharp Corp | Inspection of signal processor |
| CN101221213A (en) * | 2008-01-25 | 2008-07-16 | 湖南大学 | Analogue circuit fault diagnosis neural network method based on particle swarm algorithm |
| CN101231673A (en) * | 2008-02-02 | 2008-07-30 | 湖南大学 | Analog circuit failure diagnosis method optimized using immune ant algorithm |
| CN101251579A (en) * | 2008-03-05 | 2008-08-27 | 湖南大学 | Analog circuit failure diagnosis method based on supporting vector machine |
| CN103245907A (en) * | 2013-01-30 | 2013-08-14 | 中国人民解放军海军航空工程学院 | Artificial circuit fault diagnosis pattern sorting algorithm |
-
2016
- 2016-06-20 CN CN201610447694.2A patent/CN105866664A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH02254377A (en) * | 1989-03-29 | 1990-10-15 | Sharp Corp | Inspection of signal processor |
| CN101221213A (en) * | 2008-01-25 | 2008-07-16 | 湖南大学 | Analogue circuit fault diagnosis neural network method based on particle swarm algorithm |
| CN101231673A (en) * | 2008-02-02 | 2008-07-30 | 湖南大学 | Analog circuit failure diagnosis method optimized using immune ant algorithm |
| CN101251579A (en) * | 2008-03-05 | 2008-08-27 | 湖南大学 | Analog circuit failure diagnosis method based on supporting vector machine |
| CN103245907A (en) * | 2013-01-30 | 2013-08-14 | 中国人民解放军海军航空工程学院 | Artificial circuit fault diagnosis pattern sorting algorithm |
Non-Patent Citations (3)
| Title |
|---|
| 左磊等: "基于粒子群-支持向量机的模拟电路故障诊断", 《系统工程与电子技术》 * |
| 李璇，彭继刚，王凯歌: "基于粒子群优化的LSSVM在模拟电路故障诊断中的应用", 《贵州师范大学学报（自然科学版）》 * |
| 鲁昌华，汪涌，王玲飞: "基于PCA和SVM的模拟电路故障诊断", 《电子测量与仪器学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778694A (en) * | 2017-01-18 | 2017-05-31 | 北京工业大学 | A kind of gear transmission noises analysis method based on set empirical mode decomposition and SVMs |
| CN107202950A (en) * | 2017-05-12 | 2017-09-26 | 南京邮电大学 | Analog-circuit fault diagnosis method based on LMD approximate entropies and SVM |
| CN110096927A (en) * | 2018-01-30 | 2019-08-06 | 西安交通大学 | Contactor diagnostic method and diagnostic system based on particle group optimizing support vector machines |
| CN111461208A (en) * | 2020-03-31 | 2020-07-28 | 贵州电网有限责任公司 | Development scale prediction method and system suitable for distributed energy supply system |
| CN113487019A (en) * | 2021-07-06 | 2021-10-08 | 湖南第一师范学院 | Circuit fault diagnosis method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105866664A (en) | Intelligent fault diagnosis method for analog circuit based on amplitude frequency features | |
| CN105046279B (en) | A kind of analog circuit fault method for classifying modes | |
| CN108520357B (en) | Method and device for judging line loss abnormality reason and server | |
| CN108414923A (en) | A kind of analog-circuit fault diagnosis method based on the extraction of depth confidence network characterization | |
| CN106526400B (en) | The earth fault diagnostic method and device of DC600V power supply system of train | |
| CN106021771A (en) | Method and device for diagnosing faults | |
| WO2022067562A1 (en) | Method and device for diagnosing fault arc, and computer-readable storage medium | |
| CN102944416B (en) | Multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades | |
| CN106201871A (en) | Based on the Software Defects Predict Methods that cost-sensitive is semi-supervised | |
| CN104052612B (en) | A kind of Fault Identification of telecommunication service and the method and system of positioning | |
| CN103064008B (en) | A kind of Nolinear analog circuit soft fault diagnostic method based on Hilbert-Huang transform | |
| CN107909118A (en) | A kind of power distribution network operating mode recording sorting technique based on deep neural network | |
| CN103886328A (en) | Functional magnetic resonance image data classification method based on brain network modular structure characteristics | |
| CN103995237A (en) | Satellite power supply system online fault diagnosis method | |
| CN109633368A (en) | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA | |
| CN105572572B (en) | Analog-circuit fault diagnosis method based on WKNN-LSSVM | |
| CN110298085A (en) | Analog-circuit fault diagnosis method based on XGBoost and random forests algorithm | |
| CN109165604A (en) | The recognition methods of non-intrusion type load and its test macro based on coorinated training | |
| CN101871994B (en) | Method for diagnosing faults of analog circuit of multi-fractional order information fusion | |
| CN105137354B (en) | One kind is based on neutral net electrical fault detection method | |
| CN108535635A (en) | A kind of analog circuit intermittent fault diagnostic method based on EEMD and HMM | |
| CN105241665A (en) | Rolling bearing fault diagnosis method based on IRBFNN-AdaBoost classifier | |
| CN108052974A (en) | A kind of method for diagnosing faults, system, equipment and storage medium | |
| CN108304567A (en) | High-tension transformer regime mode identifies and data classification method and system | |
| CN109389325A (en) | Transformer substation electronic transducer state evaluating method based on wavelet neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160817 |
|
| RJ01 | Rejection of invention patent application after publication |