CN102944416B - Multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades - Google Patents

Abstract

The invention discloses a multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades. According to the method, the problems of lack of fault information and the like caused by the insufficiency of sensors is solved by adopting a plurality of sensors. An independent classifier is used for performing primary diagnosis on information acquired by each sensor, so as to determine the possibility that to-be-diagnosed faults belong to different faults; and the decision fusion diagnosis is performed by adopting a fuzzy integral fusion technology based on that the importance degree of information output by each classifier is adequately considered. According to the fault diagnosis method disclosed by the invention, classified results of all the classifiers are integrated, and the importance degree of each classifier is considered, thus effectively improving the accuracy of the fault diagnosis on the wind turbine blades.

Description

Based on the wind power generation unit blade method for diagnosing faults of multiple sensor signals integration technology

Technical field

The invention belongs to on-line monitoring and fault diagonosing technical field, especially a kind of method of the wind power generation unit blade fault diagnosis based on multiple sensor signals integration technology.

Background technology

In recent years, due to the shortage of resource and the deterioration of environment, to make countries in the world start to pay attention to development and utilization renewable and without the energy of discharge.Wind resource, as the resource of a kind of green, environmental protection, more and more obtains the attention of people.In the world, the operation of a large amount of Wind turbines makes the safe and stable operation of Wind turbines cause showing great attention to of people.Due to Wind turbines long-term work in the wild, to be exposed to the sun and in the rugged surroundings such as thunderstorm, wind field wind regime is complicated and changeable, very easily causes various fault, and therefore, the on-line monitoring and fault diagonosing of Wind turbines has become requisite link.The blade fault type of Wind turbines comprises leaf quality imbalance fault, blade aerodynamic imbalance, driftage and disconnected blade etc., because wind power generation unit blade is expensive, difficult in maintenance after damaging, therefore the condition monitoring and fault diagnosis of blade is seemed particularly important.In the initial stage Timeliness coverage fault that blade fault occurs, processed in time before problem worse affects unit operation, can greatly reduce blade maintenance, upkeep cost and difficulty.

In the process of wind power generation unit blade fault diagnosis, the data of process are all collected by sensor.Because diagnosis object operating condition is complicated, influence factor is numerous, same fault often has different performances, same symptom is usually again the coefficient result of several fault, narrowly, between detection limit and fault signature, be all a kind of Nonlinear Mapping between fault signature and the source of trouble, the fault characteristic value only relying on single-sensor to obtain generally cannot complete fault diagnosis effectively, and one of effective means solved the problem just adopts multiple sensor signals integration technology.

The mode of information fusion, generally in sensor layer, characteristic layer and decision-making level, often uses and merges in decision-making level.The information fusion technology of decision-making level is that two or more sorters is carried out integrated, adopts certain blending algorithm to diagnose.Existing blending algorithm mainly contains bayes method, D-S evidential reasoning method and fuzzy integral method.Bayes method needs prior imformation, and this prior imformation is often difficult to obtain in actual applications; And requiring that the element of decision-making set is separate, this requirement is too harsh.D-S evidential reasoning bill requires that the evidence used must be separate, is generally difficult to meet, in addition, also there will be shot array, events conflict etc.Fuzzy set can describe indeterminacy phenomenon well, and the fusion method therefore based on fuzzy integrals theory is a kind of instrument be most widely used.Fuzzy integral is a kind of non-linear Decision fusion method based on fog-density, and the classification results of integral process not only comprehensive each sorter, also considers the significance level of each sorter, fuzzy integral is applied to fault diagnosis, can reach accurate localizing faults.

Summary of the invention

The object of the invention is to overcome the deficiencies in the prior art, provide a kind of simple, cost is low, effectively can improve wind power generation unit blade safety, the wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology of reliability.

The technical solution used in the present invention is: a kind of wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology, Wind turbines installs multisensor, the failure message shortcoming problem brought because sensor is not enough is solved by adopting multisensor, sorter is adopted to carry out tentative diagnosis to the information that each sensor gathers, determine to treat that tracing trouble is under the jurisdiction of the possibility of different faults, on the basis of correlation degree taking into full account each sorter and different faults type, fuzzy integral fusion method is adopted to carry out Decision fusion diagnosis, the concrete steps of its diagnostic method are:

(1) acceleration transducer by being arranged on wind generator set main shaft seat horizontal direction and vertical direction measures blade at the vibration signal normally and under typical fault type, the corresponding sorter of each acceleration transducer;

(2) utilize empirical mode decomposition to decompose the vibration signal in the horizontal direction gathered and vertical direction respectively, different fault signatures is reacted to different intrinsic mode functions;

(3) energy of several intrinsic mode functions before calculating, the characteristic information of forming reactions fault, to extracted fault characteristic information normalized, obtains fault feature vector, as training sample and test sample book;

(4) training sample in horizontal direction and vertical direction and test sample book is utilized to carry out training and testing to two sorters respectively;

(5) fog-density and fuzzy mearue is determined; $g_j^i=g_j\left(\right\{y_i\left\}\right),i=\mathrm{1,2},\·\·\·c,j=\mathrm{1,2},\·\·\·t,$ represent the fog-density of a jth information of i-th sorter, t is the number of fault type, and c is the number of sorter; Make Y={y ₁, y ₂... y _cthe set that forms for c sorter, A _i={ y ₁, y ₂... y _i, utilize each sorter in testing to the correct recognition rata of each fault as the correlation degree of this sorter to each fault type, i.e. fog-density, then according to fog-density determination fuzzy mearue;

(6) vibration signals measured carried out empirical mode decomposition and extract fault feature vector, utilizing sorter to carry out primary fault diagnosis respectively, the result obtained is p _j=(p _j(y ₁), p _j(y ₂) ... p _j(y _c)), wherein p _j(y _i) presentation class device y _iexample to be diagnosed is divided into the possibility of jth class;

(7) utilize Choquet fuzzy integral to do fusion operator and fusion treatment is carried out to determine the fault type of wind power generation unit blade to primary fault diagnostic result; According to formula calculate fuzzy integral value e _j, e _jfor the likelihood of failure index that comprehensive diagnos goes out, then forming fault may index set E={e ₁, e ₂... e _t, according to its failure judgement type, the classification in E corresponding to maximal value is the fault type of this example.

As preferably, the concrete account form of described step (5) fuzzy mearue is:

According to formula determine λ _j, then according to formula g _j(A ₁)=g _j({ y ₁) and formula g _j(A _i)=g _j({ y _i)+g _j(A _i-1)+λ _jg _j( _yi}) g _j(A _i-1), i=1,2 ... c, asks for fuzzy mearue g _j(A _i); λ _jit is an intermediate variable.

As preferably, in described step (6), the likelihood probability that support vector machine exports is as treating in primary diagnosis that tracing trouble is under the jurisdiction of the possibility of certain fault, and it can be used as the result that primary fault is diagnosed, likelihood probability is calculated as follows

$p_j\left(y_i\right)=\left\{\begin{array}{cc}\frac{1}{1+\exp \left(\ln \left(\left(\mathrm{\ϵ}\right)f_j\left(i\right)\right)\right)}& \left|f_j\left(i\right)\right|\≤1\\ 1& \left|f_j\left(i\right)\right|>1\end{array}\right.$

In formula: f _ji () is the output of a jth support vector machine in i-th sorter; p _j(y _i) representing the likelihood probability of a jth fault in i-th sorter, likelihood probability calculating is carried out for just identifying sample class; Work as f _ji, during () >0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) > 0.5); Work as f _ji, during () <0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) <0.5); ε is arbitrarily small positive integer.

Beneficial effect: 1, the present invention adopts and extracts fault feature vector based on empirical mode decomposition, can improve the resolution of fault.

2, the present invention adopts the classification results of Fuzzy Integral Fusion not only comprehensive each sorter, also take into account the significance level of each sorter, effectively improves the accuracy of system diagnostics.

3, the present invention can carry out wind power generation unit blade localization of fault exactly, shortens maintenance and searches the time, improve the efficiency of maintenance maintenance.

4, the present invention simple, diagnosis cost low, be a kind of wind power generation unit blade method for diagnosing faults that effectively can improve wind power generation unit blade safety, reliability.

Accompanying drawing explanation

Fig. 1 is the block scheme based on the wind power generation unit blade fault diagnosis of multiple sensor signals integration technology in the present invention;

Fig. 2 is Wind turbines schematic diagram in the present invention;

Fig. 3 is empirical mode decomposition block scheme in the present invention;

Fig. 4 is based on the block scheme that the fault feature vector of empirical mode decomposition extracts in the present invention.

Embodiment

Below in conjunction with the drawings and specific embodiments, the present invention will be further described.

As shown in Figure 1, a kind of wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology, Wind turbines installs multisensor, the failure message shortcoming problem brought because sensor is not enough is solved by adopting multisensor, sorter is adopted to carry out tentative diagnosis to the information that each sensor gathers, determine to treat that tracing trouble is under the jurisdiction of the possibility of different faults, on the basis of correlation degree taking into full account each sorter and different faults type, adopt fuzzy integral fusion method to carry out Decision fusion diagnosis, concrete steps are as follows:

(1) as shown in Figure 2, the wind wheel 1 of Wind turbines is connected with main shaft, main shaft is arranged in spindle drum 2, main shaft is connected with generator 4 by shaft coupling 3, blade is measured at the vibration signal normally and under typical fault type, the corresponding sorter of each acceleration transducer by two acceleration transducers being arranged on wind generator set main shaft seat 2 horizontal direction and vertical direction.

(2) utilize empirical mode decomposition (EMD) to decompose the vibration signal in the horizontal direction gathered and vertical direction respectively, different fault signatures is reacted to different intrinsic mode functions (IMF).

Empirical mode decomposition is a kind of method processing non-stationary, nonlinear properties.Burst is decomposed into intrinsic mode functions (IMF) and a residual volume sum of multiple different frequency range by the method, and can give prominence to the local feature of signal well, its computation process as shown in Figure 3.After the calculating of series, original signal x (t) can decompose as follows

$x\left(t\right)={\mathrm{\&Sigma;}}_{i=1}^n{c}_i\left(t\right)+r_n\left(t\right)---\left(1\right)$

Therefore, any one signal x (t) can be resolved into a n IMF and residual components sum, intrinsic mode functions c ₁(t), c ₂(t), c ₃(t) ..., c _nt the composition of () respectively representation signal different frequency range from high to low, the frequency content that each frequency range comprises is different, and can change along with the change of vibration signal x (t), survival function r _nthe average tendency of (t) representation signal.

(3) energy of several IMF before calculating, the characteristic signal of forming reactions fault, to extracted fault characteristic information normalized, obtains fault feature vector, as training sample and test sample book.

Fig. 4 is the block scheme extracted based on the fault feature vector of empirical mode decomposition, and it is specially:

Step 1: carry out EMD decomposition to vibration signal sequence, obtains IMF component, and the IMF component number of different vibration signals is different, and before selecting, m IMF component is as research object.

Step 2: the energy of m IMF before calculating:

$E_i={\&Integral;}_{-\&infin;}^{+\&infin;}{\left|c_i\left(t\right)\right|}^2\mathrm{dt}$ (i=1,2…,m)(2)

Step 3: the structure of proper vector:

T＝[E ₁，E ₂，…E _m](3)

Because the energy of some IMF is comparatively large, for the ease for the treatment of and analysis, T is normalized.Proper vector is

T′＝[E ₁/E，E ₂/E，…E _m/E](4)

In formula:

(4) training sample in horizontal direction and vertical direction and test sample book is utilized to carry out training and testing to two sorters respectively.

SVM asks optimal classification surface to propose from linear separability situation.So-called optimal separating hyper plane, is exactly require that classification plane not only can be separated faultless for two class samples, and the cluster between two classes will be made maximum.For two class classification situations, its objective function is:

$\max W\left(\mathrm{\&alpha;}\right)=-\frac{1}{2}\underset{i=1}{\overset{1}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{1}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i{\mathrm{\&alpha;}}_j{y}_i{y}_{j}K(x_i,x_j)+\underset{i=1}{\overset{1}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i$

$s.t.{\mathrm{\&Sigma;}}_{i=1}^1{y}_1{\mathrm{\&alpha;}}_i=0---\left(5\right)$

0≤α _i≤C，i＝1，2，…1

And decision function is

$f\left(x\right)=\mathrm{sign}({\mathrm{\&Sigma;}}_{i=1}^1{\mathrm{\&alpha;}}_i{y}_{i}K(x_i,x)+b)---\left(6\right)$

For multicategory classification problem, adopt one-against-all Combination of Methods two class support vector machines to construct multi classifier in the present invention, two sorters adopt identical combined method.Specific as follows:

To with a t class problem, one-against-all method needs t two class support vector machines, namely t Optimal Separating Hyperplane is adopted to classify, by the tag location+1 of the i-th class, the tag location-1 of all the other all samples, then utilizes training sample and test sample book to carry out training and testing to sorter respectively.

(5) fog-density and fuzzy mearue is determined. $g_j^i=g_j\left(\right\{y_i\left\}\right),i=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;c,j=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;t,$ represent the fog-density of a jth information of i-th sorter, t is the number of fault type, and c is the number of sorter.Make Y={y ₁, y ₂... y _cthe set that forms for c sorter, A _i={ y ₁, y ₂... y _i, c is 2, utilize each sorter in testing to the correct recognition rata of each fault as the correlation degree of this sorter to each fault type, i.e. fog-density, then according to fog-density determination fuzzy mearue.

According to formula determine λ _j, then according to formula g _j(A ₁)=g _j({ y ₁) and formula g _j(A _i)=g _j({ y _i)+g _j(A _i-1)+λ _jg _j({ y _i) g _j(A _i-1), i=1,2 ... c, asks for fuzzy mearue g _j(A _i).λ _jit is an intermediate variable.

(6) vibration signals measured carried out empirical mode decomposition and extract fault feature vector, utilizing sorter to carry out primary fault diagnosis respectively, the result obtained is p _j=(p _j(y ₁), p _j(y ₂) ... p _j(y _c)), wherein p _j(y _i) presentation class device y _iexample to be diagnosed is divided into the possibility of jth class, wherein

$p_j\left(y_i\right)=\left\{\begin{array}{cc}\frac{1}{1+\exp \left(\ln \left(\left(\mathrm{\&epsiv;}\right)f_j\left(i\right)\right)\right)}& \left|f_j\left(i\right)\right|\&le;1\\ 1& \left|f_j\left(i\right)\right|>1\end{array}\right.---\left(7\right)$

In formula: f _ji () is the output of a jth support vector machine in i-th sorter; p _j(y _i) representing the likelihood probability of a jth fault in i-th sorter, likelihood probability calculating is carried out for just identifying sample class.Work as f _ji, during () > 0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) > 0.5); Work as f _ji, during () <0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) <0.5); ε is arbitrarily small positive integer.

(7) utilize Choquet fuzzy integral to do fusion operator and fusion treatment is carried out to determine the fault type of wind power generation unit blade to primary fault diagnostic result.According to formula calculate fuzzy integral value e _i, e _jfor the likelihood of failure index that comprehensive diagnos goes out, then forming fault may index set E={e ₁, e ₂... e _t, according to its failure judgement type, the classification in E corresponding to maximal value is the fault type of this example.

It should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.The all available prior art of each ingredient not clear and definite in the present embodiment is realized.

Claims (3)

1. the wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology, it is characterized in that: on Wind turbines, multisensor is installed, sorter is adopted to carry out tentative diagnosis to the information that each sensor gathers, determine to treat that tracing trouble is under the jurisdiction of the possibility of different faults, on the basis of correlation degree taking into full account each sorter and different faults type, adopt fuzzy integral fusion method to carry out Decision fusion diagnosis, the concrete steps of its diagnostic method are:

(1) acceleration transducer by being arranged on wind generator set main shaft seat horizontal direction and vertical direction measures blade at the vibration signal normally and under typical fault type, the corresponding sorter of each acceleration transducer;

(2) utilize empirical mode decomposition to decompose the vibration signal in the horizontal direction gathered and vertical direction respectively, different fault signatures is reacted to different intrinsic mode functions;

(3) energy of several intrinsic mode functions before calculating, the characteristic information of forming reactions fault, to extracted fault characteristic information normalized, obtains fault feature vector, as training sample and test sample book;

(4) training sample in horizontal direction and vertical direction and test sample book is utilized to carry out training and testing to two sorters respectively;

(5) fog-density and fuzzy mearue is determined; $g_j^i=g_j\left(\right\{y_i\left\}\right),i=\mathrm{1,2},...c,j=\mathrm{1,2},....t,$ represent the fog-density of a jth information of i-th sorter, t is the number of fault type, and c is the number of sorter; Make Y={y ₁, y ₂... y _cthe set that forms for c sorter, A _i={ y ₁, y ₂... y _i, utilize each sorter in testing to the correct recognition rata of each fault as the correlation degree of this sorter to each fault type, i.e. fog-density, then according to fog-density determination fuzzy mearue;

(6) vibration signals measured carried out empirical mode decomposition and extract fault feature vector, utilizing sorter to carry out primary fault diagnosis respectively, the result obtained is p _j=(p _j(y ₁), p _j(y ₂) ... p _j(y _c)), wherein p _j(y _i) presentation class device y _iexample to be diagnosed is divided into the possibility of jth class;

(7) utilize Choquet fuzzy integral to do fusion operator and fusion treatment is carried out to determine the fault type of wind power generation unit blade to primary fault diagnostic result; According to formula calculate fuzzy integral value e _j, e _jfor the likelihood of failure index that comprehensive diagnos goes out, then forming fault may index set E={e ₁, e ₂... e _t, according to its failure judgement type, the classification in E corresponding to maximal value is the fault type of this example.

2. the wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology according to claim 1, is characterized in that: the concrete account form of described step (5) fuzzy mearue is:

According to formula determine λ _j, then according to formula g _j(A ₁)=g _j({ y ₁) and formula g _j(A _i)=g _j({ y _i)+g _j(A _i-1)+λ _jg _j({ y _i) g _j(A _i-1), i=1,2 ... c, asks for fuzzy mearue g _j(A _i); λ _jit is an intermediate variable.

3. the wind power generation unit blade method for diagnosing faults based on multiple sensor signals integration technology according to claim 1, it is characterized in that: in described step (6), the likelihood probability that support vector machine exports is as treating in primary diagnosis that tracing trouble is under the jurisdiction of the possibility of certain fault, it can be used as the result that primary fault is diagnosed, likelihood probability is calculated as follows

$p_j\left(y_i\right)=\left\{\begin{array}{cc}\frac{1}{1+\exp \left(\ln \left(\mathrm{\&epsiv;}\right)f_j\left(i\right)\right))}& \left|f_j\left(i\right)\right|\&le;1\\ 1& \left|f_j\left(i\right)\right|>1\end{array}\right.$

In formula: f _ji () is the output of a jth support vector machine in i-th sorter; p _j(y _i) representing the likelihood probability of a jth fault in i-th sorter, likelihood probability calculating is carried out for just identifying sample class; Work as f _ji, during () >0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) >0.5); Work as f _ji, during () <0, it is p that sample is assigned to the degree of membership just identifying class _j(y _i) (p _j(y _i) <0.5); ε is arbitrarily small positive integer.