CN102129027A - Fault diagnosis method for switched-current circuit based on fault dictionary - Google Patents

Abstract

The invention discloses a fault diagnosis method for a switched-current circuit based on a fault dictionary. The fault diagnosis method comprises the following steps of: 1) acquiring an electrical signal of the switched-current circuit, wherein the electrical signal is a testable node time domain response signal of the switched-current circuit; 2) centralizing the acquired electrical signal; 3) defining a circuit fault mode; 4) calculating a signal entropy of the electrical signal; 5) finding out a fuzzy set of the signal entropy; and 6) establishing the fault dictionary according to the signal entropy and the fuzzy set of the signal entropy, and classifying the various faults of the tested circuit by using the fault dictionary. By the fault diagnosis method, a hard fault can be classified and a soft fault can be positioned; at the same time, the fault diagnosis method can be applied to both a real time system and a non-real time system and has a wide application range.

Description

A kind of Switched-Current Circuit method for diagnosing faults based on fault dictionary

Technical field

The invention belongs to pattern-recognition and electronic circuit engineering field, be specifically related to a kind of Switched-Current Circuit method for diagnosing faults based on fault dictionary.

Background technology

Traditional simulation and Switched-Current Circuit fault comprise bust (hard fault) and parameter fault (soft fault), the open circuit of bust finger element and short circuit failure of removal, circuit performance is caused great influence, this fault can change the topological structure of circuit, and diagnosis is got up relatively easy.And the parameter fault is meant that the parameter of circuit component exceeds predetermined range of tolerable variance, and this class fault can not change the topological structure of circuit, and generally they do not make equipment complete failure, but can cause circuit performance seriously to descend.

About switch current circuit testing and fault diagnosis aspect, because components and parts exist tolerance, nonlinear problem and relatively poor factors such as fault model, Switched-Current Circuit and simulation system fault diagnosis difficulty are increased greatly, it is become have challenging research topic.In general, there is skew at random in component parameters in range of tolerable variance, because the ubiquity of tolerance causes fault to have ambiguity, i.e. the measurability variation of fault.From the practice of circuit fault diagnosis, the greatest difficulty that fault diagnosis faces is the tolerance of component parameters.Actual simulation and Switched-Current Circuit often contain nonlinear element, and the also existence often in linear circuit of a large amount of nonlinear problems, so when fault diagnosis, have huge calculated amount and complicated computing formula unavoidably.In simulation and the Switched-Current Circuit, the relation between input and the output is complicated and be difficult to modelling, causes still lacking valid model so far to general diagnosing malfunction.

In all multi-methods of Switched-Current Circuit fault diagnosis, fault dictionary method is highly suitable for diagnosing non-linear circuit and on-line fault diagnosis because diagnosable condition does not have strict restriction, and practicality is stronger, more and more comes into one's own.But, traditional fault dictionary method generally only is applicable to the diagnosis of single hard fault, for soft fault, because the influence of the tolerance in the circuit lacks system and effective disposal route, so traditional fault dictionary method is not too suitable, there is bigger limitation in traditional fault dictionary method.

Summary of the invention

In order to overcome the problems referred to above that prior art exists, the invention provides a kind of Switched-Current Circuit method for diagnosing faults based on fault dictionary applied widely.

Technical scheme of the present invention is: it may further comprise the steps:

1) electric signal of collection Switched-Current Circuit, but described electric signal is the test node time domain response signal of Switched-Current Circuit;

2) electric signal of gathering being done centralization handles;

3) definition circuit fault mode;

4) the signal entropy of calculating electric signal;

5) find the fuzzy set of signal entropy;

6) set up fault dictionary according to the fuzzy set of signal entropy and signal entropy, utilize fault dictionary being classified by the various faults of diagnostic circuit.

The method of gathering the Switched-Current Circuit electric signal in the described step 1) is: but extract the time domain response signal with data acquisition board at the side circuit test node, or the various malfunctions of circuit are carried out emulation with ASIZ software, collect circuit time domain response signal.

Described step 2) in the electric signal of gathering being done the method that centralization handles is: signal vector x carries out centralization, i.e. x ← x-E{x} with the method that deducts average.

The method of definition circuit fault mode is in the described step 3): to diagnose six rank chebyshev low-pass filter soft faults is example, supposes that the transistor that breaks down simultaneously in the circuit has 5, is respectively Mg1, Mf1, Me2, Md1 and Mj.When transistor Mg1 transconductance value is higher or lower than its nominal value 50%, and other four metal-oxide-semiconductors change in its range of tolerable variance, at this moment resulting fault mode be respectively Mg1 ↑ and Mg1 ↓.Described six rank chebyshev low-pass filter soft fault patterns comprise Mg1 ↑, Mg1 ↓, Mf1 ↑, Mf1 ↓, Me2 ↑, Me2 ↓, Md1 ↑, Md1 ↓, Mj ↑, Mj ↓.

The computing formula of calculating the signal entropy of electric signal in the described step 4) is:

J(x)≈k ₁(E{G ¹(x)}) ²+k ₂(E{G ²(x)}-E{G ²(v)}) ²；

In the formula,

k ₁And k ₂Be positive constant, they are the stochastic variables that satisfy standardized normal distribution; E represents expectation value; G ¹Be odd function, G ²Be even function, x is the variable of electric signal correspondence, and v is certain stochastic variable, and its probability density function is the Gaussian distribution of standard, Wherein

As a simple special case, described G ¹(x)=xexp (x ²/ 2), G ²(x)=| x|.

The method of the fuzzy set that finds the signal entropy in the described step 5) is: when the transistor transconductance value in the Switched-Current Circuit changes in range of tolerable variance, corresponding each fault mode, moving 50 Monte Carlos (Monto-Carlo) analyzes, produce 50 time domain responses, find the signal entropy fuzzy set.

Fault dictionary in the described step 6) comprises soft fault class fault dictionary and hard fault class fault dictionary, and soft fault class fault dictionary structure comprises: the fuzzy set of fault mode, failure code, nominal value, fault value, actual signal entropy, ideal signal entropy and signal entropy; Hard fault class fault dictionary structure comprises: the fuzzy set of fault mode, failure code, actual signal entropy, ideal signal entropy and signal entropy.

Failure code is the pairing code of each fault mode, for example available F0, and F1, F2 ... represent; Nominal value is the normalization transconductance value of MOS transistor in the Switched-Current Circuit; Fault value is skew MOS transistor nominal value 50% a resulting transistor transconductance value; But the actual signal entropy be according to data acquisition board in the time domain response data that the side circuit test node collects, be updated to again in the signal entropy formula and calculate resulting value; The ideal signal entropy is to obtain the time domain response data according to the ASIZ software emulation, is updated to signal entropy formula again and calculates resulting value.

The diagnosis principle of described fault dictionary comprises two courses of work: make up fault dictionary process (learning process) and failure diagnostic process (test process).

The present invention uses above-mentioned principle, at first utilize data acquisition board to reality by diagnostic circuit (CUTs) but test node extracts the time domain response signal, perhaps utilize ASIZ software that circuit is carried out emulation, from the time domain output response signal of circuit, extract raw data, in MATLAB software, calculate its signal entropy then, raw data is carried out pre-service (being the signal calculated entropy), like this, extracted the fault signature parameter under every kind of fault mode, and find the fuzzy set of signal entropy, make up a fault dictionary.Fault dictionary will be to being classified by the various faults of diagnostic circuit.

Practice shows, the transistor transconductance value G in Switched-Current Circuit _mWhen changing in 5% or 10% range of tolerable variance, the signal entropy under the different faults pattern drops on different zones, therefore, when response is not overlapping, can correctly detect defective transistor.Though consider tolerance some overlapped data are arranged, more than the Fault Identification rate of accuracy reached to 95%.

Use the present invention, when the signal calculated entropy, the number of the fault signature parameter of extracting is few, and a unique characteristic parameter is only arranged, and has reduced calculating and failure diagnosis time; The small scale of constructed fault dictionary, simple in structure, failure detection time is short, and the robotization that is easy to the system that realizes is handled; Use the present invention, the hard fault of can not only classifying also can be located soft fault; Simultaneously, also can be used in real time and the non real-time system, applied widely.

Description of drawings

Fig. 1 is a fault diagnosis principle block diagram of the present invention.

Process shown in the dotted arrow is for making up the fault dictionary process among the figure, and process shown in the solid arrow is fault test and diagnostic procedure.

Embodiment

The present invention will be described in detail below in conjunction with drawings and Examples.

With reference to Fig. 1, present embodiment comprises the steps:

1) electric signal of collection Switched-Current Circuit, but described electric signal is the test node time domain response signal of Switched-Current Circuit; Concrete grammar is: with data acquisition board by diagnostic circuit (CUTs) but test node extracts the time domain response signal, or the various malfunctions of circuit are carried out emulation with ASIZ software, collect circuit time domain response signal.

2) electric signal of gathering is done centralization and handle, signal vector x is carried out centralization, i.e. x ← x-E{x} with the method that deducts average;

3) definition circuit fault mode;

4) the signal entropy of calculating electric signal;

5) find the fuzzy set of signal entropy;

6) set up fault dictionary according to signal entropy and signal entropy fuzzy set, utilize fault dictionary being classified by the various faults of diagnostic circuit.

Described step 2) in, the concrete grammar of the electric signal of gathering being done the centralization processing is: signal vector x is carried out centralization, i.e. x ← x-E{x} with the method that deducts average;

This average E{x} is by data acquisition sample x (1) in practice, x (2) ..., x (n) calculates that its expectation obtains.

In the described step 3), the method for definition circuit fault mode is: to diagnose six rank chebyshev low-pass filter soft faults is example, supposes that the transistor that breaks down simultaneously in the circuit has 5, is respectively Mg1, Mf1, Me2, Md1 and Mj.When transistor Mg1 transconductance value is higher or lower than its nominal value 50%, and other four metal-oxide-semiconductors change in its range of tolerable variance, at this moment resulting fault mode be respectively Mg1 ↑ and Mg1 ↓.Described six rank chebyshev low-pass filter soft fault patterns comprise Mg1 ↑, Mg1 ↓, Mf1 ↑, Mf1 ↓, Me2 ↑, Me2 ↓, Md1 ↑, Md1 ↓, Mj ↑, Mj ↓.

The signal entropy is finished by step 4).For the stochastic variable X of a discrete value, its entropy H is defined as:

$H\left(X\right)=-\underset{i}{\mathrm{\Σ}}p(X=a_i)\log p(X=a_i)---\left(1\right)$

In the formula, a _iBe the possible value of X, P (X=a _i) be X=a _iProbability density.Logarithm is got different substrates, will obtain the not commensurate of entropy.Usually use 2 as substrate, unit is called bit in this case.

Function f is defined as:

f(p)＝-p?log?p?0≤p≤1 (2)

This is a non-negative function, for just, utilizes this function when p gets intermediate value, can be write as entropy:

$H\left(X\right)=\underset{i}{\mathrm{\Σ}}f\left(p(X=a_i)\right)---\left(3\right)$

For continuous signal, the computing formula of its differential entropy H (x) is:

H(x)＝-∫p _x(ξ)logp _x(ξ)dξ＝∫f(p _x(ξ))dξ(4)

Here p (x) is the probability density function of signal x,

Suppose that we have estimated n the different function F of x ⁱ(x) expectation c _i:

E{F ⁱ(x)}＝∫p(x)F ⁱ(x)dx＝c _i，i＝1，2，…，n (5)

Maximum entropy theorem shows, under suitable systematicness condition, satisfies constraint formula (5), and have the density p of very big entropy in all this density _o(ξ), shape as:

$p\left(x\right)=\mathrm{Aexp}\left(\underset{i}{\mathrm{\Σ}}{a}_i{F}^i\left(x\right)\right),i=1,\·\·\·,n---\left(6\right)$

In the formula, A and a _iBe the constraint that utilizes in the formula (5), be about to the p in formula (6) the right alternate form (5), and constraint ∫ p _o(ξ) d ξ=1 is from c _iThe constant of determining.

Now, consideration can be got all values on the number line, has zero-mean and certain constant variance the set (so we have two constraints) of the stochastic variable of (for example 1).For this kind variable, it is Gaussian distribution that its very big entropy distributes.By formula (6), this density has following form:

$p_o\left(\mathrm{\ξ}\right)=\mathrm{Aexp}(\underset{i}{\mathrm{\Σ}}{a}_1{\mathrm{\ξ}}^2+a_2\mathrm{\ξ})---\left(7\right)$

Following formula shows that gaussian variable has very big entropy in having all stochastic variables of unit variance.This means that entropy can be used as a kind of tolerance of non-Gauss.

F ⁱ(x) can be any one group of linear function about x.Again because ∫ p _o(ξ) d ξ=1, so a total n+1 nonlinear equation need be found the solution, this needs the method for numerical evaluation usually, and is difficult to finish.We will introduce approaching of entropy based on approximate maximum entropy method.The motivation of the method is such: the entropy of a distribution can not just can be determined by limited expectation, even they all estimate very accurately as in formula (5).As explaining in front, there is infinite a plurality of distribution, they all satisfy the constraint in the formula (5), but their entropy differs greatly.Especially, only get as x under the limit situations of limited value, differential entropy is tending towards-∞.A simple solution is a maximum entropy method.This means that what we calculated is very big entropy, it with we constraint formula (5) or observe and can compare, and this is a well posed problem.This very big entropy, perhaps further approaching of it can be used as the significant of an entropy of a random variable and approach.We will be under some given constraints, and the single order of at first deriving the very big entropy density of a stochastic variable continuous, one dimension approaches.Note now, mean every other a in the formula (6) near Gauss's hypothesis _iWith a _N+2≈-1/2 compares very little, because the exponential sum exp (ξ in the formula (6) ²/ 2) very close.Like this, our single order that can get exponential function approaches.Can obtain simply separating of the middle constant of formula (6) thus, and we have obtained approximate very big entropy density, we are designated as it

In the formula, c _i=E{F ⁱ(ξ) }.Now, this that utilizes density is approximate, and one that can derive differential entropy is approached:

$J\left(x\right)\≈\frac{1}{2}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}E{\left\{F^i\left(x\right)\right\}}^2---\left(9\right)$

Now, only remaining " tolerance " function F of selecting information in the definition (5) ⁱ.Its concrete implementation process is to select any one group of linearity independently function, for example G earlier ⁱ, i=1 ..., m, and then to comprising these functions and monomial ξ ^k, k=0, Gram-Schmidt orthonomalization, the collection of functions F that obtains are used in 1,2 set ⁱSatisfy the orthogonality hypothesis.When the actual selection function, should emphasize following criterion:

(1) E{G ⁱ(x) } actual estimated should not had any problem on statistics.Especially, this estimates should not be worth the open country too responsive;

(2) in order to guarantee the existence of very big entropy, G ⁱ(x) growth should be not faster than quadratic function;

(3) G ⁱ(x) must catch those parts that when calculating its entropy, are concerned with in the distribution of x.

Those top criterions only limit operable function space.Our framework allows to use different functions as G ⁱ

If we use two function G ¹And G ², their selection makes G ¹Be odd function and G ²Be even function, just obtain a kind of special case of formula (8).Odd function has been measured skew-symmetry, and even function has been measured the size of zero place's bimodal relative peak, and this is more closely related with time this property of the relative superelevation of Gauss.At this in particular cases, the approximate maximum entropy of the signal in the formula (9) is approximate is reduced to:

J(x)≈k ₁(E{G ¹(x)}) ²+k ₂(E{G ²(x)}-E{G ²(v)}) ² (10)

Wherein, k ₁And k ₂It is positive constant.It is the stochastic variable that satisfies standardized normal distribution.Here E (x) represents the expectation value of variable x, G ¹And G ²Satisfy above three rules, and G ¹Be odd function, G ²Be even function.The v here is with the v unanimity that defines in the top formula, promptly

Wherein

For example, when selecting G ¹(x)=x exp (x ²/ 2), G ²(x)=| during x|, maximum entropy calculates and is approximately:

$J\left(x\right)=k_1{\left(E\right\{\mathrm{xexp}(-x^2/2)\left\}\right)}^2+k_2{\left(E\right\{\left|x\right|\}-\sqrt{2/\mathrm{\π}})}^2---\left(11\right)$

Wherein, $k_1=36/(8\sqrt{3}-9),k_2=1/(2-6/\mathrm{\π})$

The method that finds signal entropy fuzzy set in the described step 5) is: when the transistor transconductance value of Switched-Current Circuit changes in range of tolerable variance, corresponding each fault mode, moving 50 Monte Carlos (Monto-Carlo) analyzes, produce 50 time domain responses, find the fuzzy set of signal entropy.With six rank chebyshev low-pass filters is example,

The fault dictionary diagnosis principle comprises two courses of work: make up fault dictionary process (learning process) and failure diagnostic process (test process).

Fault dictionary in the described step 6) comprises soft fault class fault dictionary and hard fault class fault dictionary, table 1 is the soft fault class fault dictionary of six rank chebyshev low-pass filters, soft fault class fault dictionary structure has 7 row, is respectively the fuzzy set of fault mode, failure code, nominal value, fault value, actual signal entropy, ideal signal entropy and signal entropy.The line number of fault dictionary is the kind of fault mode.

The soft fault class fault dictionary of table 1 six rank chebyshev low-pass filters

Table 2 is the hard fault class fault dictionary of six rank chebyshev low-pass filters, and hard fault class fault dictionary structure has 5 row, the fuzzy set of fault mode, failure code, actual signal entropy, ideal signal entropy and signal entropy.The line number of fault dictionary is the kind of fault mode.

The hard fault class fault dictionary of table 2 six rank chebyshev low-pass filters

Claims (7)

1. the Switched-Current Circuit method for diagnosing faults based on fault dictionary is characterized in that, may further comprise the steps:

1) electric signal of collection Switched-Current Circuit, but described electric signal is the test node time domain response signal of Switched-Current Circuit;

2) electric signal of gathering being done centralization handles;

3) definition circuit fault mode;

4) the signal entropy of calculating electric signal;

5) find the fuzzy set of signal entropy;

6) set up fault dictionary according to the fuzzy set of signal entropy and signal entropy, utilize fault dictionary being classified by the various faults of diagnostic circuit.

2. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 1, it is characterized in that, the method of gathering the Switched-Current Circuit electric signal in the described step 1) is: but extract the time domain response signal with data acquisition board at the side circuit test node, or the various malfunctions of circuit are carried out emulation with ASIZ software, collect circuit time domain response signal.

3. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 1 and 2, it is characterized in that, described step 2) in the electric signal of gathering being done the method that centralization handles is: signal vector x carries out centralization, i.e. x ← x-E{x} with the method that deducts average.

4. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 1 and 2 is characterized in that, the computing formula of calculating the signal entropy of electric signal in the described step 4) is:

J(x)≈k ₁(E{G ¹(x)}) ²+k ₂(E{G ²(x)}-E{G ²(v)}) ²；

In the formula, k ₂=1/ (2-6/ π), k ₁And k ₂Be positive constant, they are the stochastic variables that satisfy standardized normal distribution; E represents expectation value; G ¹Be odd function, G ²Be even function, x is the variable of electric signal correspondence, and v is certain stochastic variable, and its probability density function is the Gaussian distribution of standard,

Wherein

5. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 5 is characterized in that described G ¹(x)=x exp (x ²/ 2), G ²(x)=| x|.

6. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 1 and 2, it is characterized in that, the method of the fuzzy set that finds the signal entropy in the described step 5) is: when changing in the transistor transconductance value range of tolerable variance in the Switched-Current Circuit, corresponding each fault mode, move 50 Monte Carlo Analysis, produce 50 time domain responses, find the fuzzy set of signal entropy.

7. the Switched-Current Circuit method for diagnosing faults based on fault dictionary according to claim 1 and 2, it is characterized in that, fault dictionary in the described step 6) comprises soft fault class fault dictionary and hard fault class fault dictionary, and soft fault class fault dictionary structure comprises: the fuzzy set of fault mode, failure code, nominal value, fault value, actual signal entropy, ideal signal entropy and signal entropy; Hard fault class fault dictionary structure comprises: the fuzzy set of fault mode, failure code, actual signal entropy, ideal signal entropy and signal entropy.

Images