CN103927752A - Method for monitoring operating state of electrical equipment - Google Patents

Abstract

The invention discloses a method for monitoring the operating state of electrical equipment. The method includes the step of acquisition of an infrared thermogram of the operating state of the electrical equipment; the step of wavelet transform, wherein wavelet transform is carried out on the infrared thermogram to extract low-frequency components and high-frequency components of the infrared thermogram; the step of homomorphic filtering, wherein homomorphic filtering is carried out on the low-frequency components; the step of fuzzy enhancement, wherein fuzzy enhancement is carried out on the high-frequency components; the step of wavelet reconstruction, wherein wavelet reconstruction is carried out on the low-frequency components having undergone homomorphic filtering and the high-frequency components having undergone fuzzy enhancement to obtain an enhanced infrared thermogram; the operating state of the electrical equipment is monitored according to the obtained enhanced infrared thermogram. According to the method for monitoring the operating state of the electrical equipment, the infrared thermogram of the electrical equipment becomes clear and accurate by being enhanced, and therefore the operating state of the electrical equipment can be better monitored.

Description

A kind of method of monitoring power equipment running status

Technical field

The present invention relates to a kind of power equipment method for monitoring operation states, relate in particular to the method for utilizing Infrared Thermogram to monitor power equipment running status.

Background technology

According to statistics, nearly 90% substation accident is caused by electrical equipment fault, and about 50% faulty equipment has thermal anomaly phenomenon in early days, therefore utilize infrared thermal imaging technique to carry out infrared thermal imaging to electric equipments, the accident defect that may exist by analyzing the variation judgement electrical equipment of Infrared Thermogram is widely applied in the band electro-detection of transformer station.But due to infrared thermal imaging, be subject to the impact of the devices such as environment and optical electron, cause that Infrared Thermogram visual effect is fuzzy, resolution and contrast lower, restricting to a certain extent the accuracy of fault diagnosis.Therefore, the Infrared Thermogram enhancing technology of improving infrared thermal imagery plot quality by raising contrast, enhancing visual effect has been subject to extensive attention, also has important construction value.

Conventional Infrared Thermogram strengthens no matter technology is frequency domain treatment technology or spatial processing technology, in the contradictory problems solving between integral body, part and efficiency, all has certain deficiency.In recent years, the image processing techniques of processing based on wavelet transformation, homomorphic filtering and fuzzy enhancing is widely applied, obtained certain effect, but choosing of fuzzy parameter all got general value or manually chosen by rule of thumb, be difficult to obtain optimized analysis effect, cause picture contrast poor, visual effect is fuzzy, is unfavorable for the diagnosis location of equipment failure.

Summary of the invention

The object of the present invention is to provide a kind of method of monitoring power equipment running status, the method has obtained clear, Infrared Thermogram accurately by the Infrared Thermogram of power equipment is strengthened, thereby can better monitor the running status of power equipment.

In the technical program, the Infrared Thermogram of monitoring power equipment running status is being strengthened in the step of processing, adopted the method for fuzzy enhancing, it is preferentially chosen automatically to fuzzy parameter, to obtain optimized analysis effect, thereby further strengthen visual effect, be beneficial to the diagnosis location of equipment failure.

To achieve these goals, the present invention proposes a kind of method of monitoring power equipment running status, it comprises the following steps:

Obtain the Infrared Thermogram of power equipment running status;

Wavelet transformation: described Infrared Thermogram is carried out to wavelet transformation, extract its low frequency component and high fdrequency component;

Homomorphic filtering: described low frequency component is carried out to homomorphic filtering;

Fuzzy enhancing step: described high fdrequency component is carried out to fuzzy enhancing;

Wavelet reconstruction: the high fdrequency component that the low frequency component that homomorphic filtering was processed and fuzzy enhancing were processed is carried out wavelet inverse transformation, the Infrared Thermogram being enhanced;

According to the running status of the Infrared Thermogram monitoring power equipment of the enhancing obtaining.

The method of monitoring power equipment running status of the present invention, by Infrared Thermogram being strengthened to processing, is intended to obtain Infrared Thermogram more clearly, thereby strengthens visual effect, carries out the diagnosis location of equipment failure.

Further, the method for monitoring power equipment running status of the present invention, in described fuzzy enhancing step, adopts dynamic self-adapting genetic algorithm to determine the required fuzzy parameter of described fuzzy enhancing.

Use self-adapted genetic algorithm, can carry out automatic optimal operation to the fuzzy parameter in fuzzy enhancing step, solved the automatic selection of optimal problem of fuzzy parameter, thereby further strengthened the visual effect of Infrared Thermogram.Specifically, former Infrared Thermogram generates low frequency component and high fdrequency component through wavelet transformation; The region that in low frequency component correspondence image, gray-scale value slowly changes, utilizes homomorphic filtering to process; In high fdrequency component correspondence image, gray-value variation is very fast, edge and detailed information be concentrated region mainly, utilize dynamic self-adapting genetic algorithm to determine that the method for fuzzy parameter carries out fuzzy enhancing processing, the result of homomorphic filtering and fuzzy enhancing processing is carried out to wavelet reconstruction, thus the Infrared Thermogram being enhanced.

Further, in the method for monitoring power equipment running status of the present invention, adopt dynamic self-adapting genetic algorithm and take gradation of image information entropy and determine the required fuzzy parameter of described fuzzy enhancing as evaluation criterion.

So-called information entropy, is exactly average information, the size of image information entropy reflected image inclusion information amount number, entropy is larger, the quantity of information that key diagram looks like to carry is larger, thereby effect is also relatively better.For arbitrary gray scale xi of gradation of image collection X, if the probability that xi occurs is P (xi), the gradation of image information entropy of gradation of image collection X is

$H\left(x\right)=-\underset{i=1}{\overset{n}{\mathrm{\Σ}}}P\left(x_i\right)\log \left(P\left(x_i\right)\right)---\left(9\right)$

In formula, P (xi) is that gray-scale value is the pixel count N (xi) of xi and the ratio of image total pixel number N, i.e. P (xi)=N (xi)/N.

Further, in the method for monitoring power equipment running status of the present invention, described dynamic self-adapting genetic algorithm coincidence formula:

$P_c=\left\{\begin{array}{c}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.---\left(3\right)$

$P_m=\left\{\begin{array}{c}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f)}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},f<f_{\mathrm{avg}}\end{array}\right.---\left(4\right)$

Wherein, P _cfor crossover probability; P _mfor variation probability; f _maxmaximum adaptation degree in Wei Meidai colony; f _avgthe average fitness of Wei Meidai colony; F ' is larger fitness in two individualities that will intersect; F is for making a variation individual fitness; P _c1, P _c2, P _m1and P _m2be a constant in interval (0,1).

Preferably, the above-mentioned Infrared Thermogram to monitoring power equipment running status strengthens in the method for processing, P _c1=0.9, P _c2=0.6, P _m1=0.1, P _m2=0.001.

In view of changeless crossover probability P in standard genetic algorithm _cwith variation probability P _mlimited search speed and the search effect of algorithm, Srinivas has proposed Adaptive Genetic (AGA) algorithm, according to the height of chromosome fitness, at population average fitness f _avgwith maximum adaptation degree f _maxbetween to P _cand P _madjust, as shown in formula (1), (2):

$P_c=\left\{\begin{array}{c}\frac{P_{c1}(f_{\max }-f^{\&prime;})}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c2},f^{\&prime;}<f_{\mathrm{avg}}\end{array}---\left(1\right)\right.$ $P_m=\left\{\begin{array}{c}\frac{P_{m1}(f_{\max }-f^{\&prime;})}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m2},f<f_{\mathrm{avg}}\end{array}---\left(2\right)\right.$

Wherein, f ' is higher fitness value in two parent chromosomes to be intersected; F is the fitness value for the treatment of mutated chromosome; P _c1, P _c2, P _m1and P _m2be a constant in interval (0,1).

As shown in formula (3), (4), the self-adapted genetic algorithm that such scheme of the present invention adopts is dynamic self-adapting genetic algorithm, wherein crossover probability P _cwith variation probability P _mautomatically to change with fitness individual in population.When the fitness value of individuality is lower than average fitness value in population, illustrate that this individuality is the bad individuality of performance, just should adopt larger crossover probability P _cwith variation probability P _m; When the fitness value of individuality is higher than average fitness value in population, this individuality function admirable is described, just should selects corresponding crossover probability P according to its fitness _cwith variation probability P _m.Adopted after dynamic self-adapting technology the crossover probability P in self-adapted genetic algorithm _cwith variation probability P _mjust can suitably automatically adjust, provide each to separate corresponding best crossover probability P _cwith variation probability P _m, so both guaranteed the diversity of population in genetic algorithm, avoid precocious, also can guarantee the convergence of genetic algorithm.

Fuzzy enhancing step of the present invention, it carries out fuzzy enhancing processing to high fdrequency component.According to fuzzy subset's concept, a width size, for the Infrared Thermogram that M * N tieed up and had K gray shade scale, can represent with a fuzzy point set matrix X

$X=\underset{i=1}{\overset{M}{\mathrm{\&cup;}}}\underset{j=1}{\overset{N}{\mathrm{\&cup;}}}(p_{\mathrm{ij}}/x_{\mathrm{ij}})---\left(5\right)$

In formula: x _ijfor image, i is capable, the gray-scale value of j row pixel; p _ijit is the membership function of (i, j) individual pixel; p _ij/ x _ijin representative image, (i, j) individual pixel gray-scale value is with respect to the brightness level of certain standard grayscale, and the selective basis actual needs of standard grayscale is determined.It generally solves suc as formula shown in (6):

$p_{\mathrm{ij}}=G\left(x_{\mathrm{ij}}\right)=[1+\frac{(x_{\max }-x_{\mathrm{ij}})}{F_d}{]}^{-\mathrm{Fe}}---\left(6\right)$

In formula: x _maxfor gradation of image maximal value; F _dand F _ebe respectively type reciprocal and the exponential type ambiguity factor (being fuzzy parameter).

At fuzzy space, strengthen conversion, as shown in formula (7):

$I\left(p_{\mathrm{ij}}\right)=\left\{\begin{array}{c}2{\left(p_{\mathrm{ij}}\right)}^2,0\&le;P_{\mathrm{ij}}\&le;0.5\\ 1-2{(1-p_{\mathrm{ij}})}^2,0.5\&le;p_{\mathrm{ij}}\&le;1\end{array}\right.---\left(7\right)$

To pixel degree of membership p _ijbe r time and strengthen conversion, the fuzzy set p after being converted _ij', as shown in formula (8), I wherein _rfor calling for r time of I.

p _ij'＝I _r(p _ij)＝I(I _r-1(p _ij)) (8)

Can effectively the stretch tonal range of image of fuzzy enhancement algorithm, increases the contrast of target image and background.

The method of monitoring power equipment running status disclosed by the invention has obtained clear, Infrared Thermogram accurately by the Infrared Thermogram of power equipment is strengthened, thereby can better monitor the running status of power equipment.

Accompanying drawing explanation

Fig. 1 is that the method for monitoring power equipment running status of the present invention strengthens to Infrared Thermogram the process flow diagram of processing under a kind of embodiment.

Fig. 2 is that the method for monitoring power equipment running status of the present invention is carried out the schematic diagram of wavelet transformation under a kind of embodiment.

Embodiment

Below in conjunction with Figure of description and specific embodiment, the method for monitoring power equipment running status of the present invention is made to further explanation and explanation.

Fig. 1 has illustrated the method for monitoring power equipment running status of the present invention, under a kind of embodiment, Infrared Thermogram is strengthened to the flow process of processing.Fig. 2 has illustrated the small wave converting method in this flow process.

As shown in Figure 1, in the present embodiment, the method for monitoring power equipment running status comprises the following steps:

Step 1: the Infrared Thermogram that obtains power equipment running status

Step 2: the Infrared Thermogram to monitoring substation electric equipment operation state carries out wavelet transformation, this wavelet transformation adopts orthogonal wavelet to decompose, in Matlab, call orthogonal wavelet transformation function to former Infrared Thermogram f (x, y) carry out orthogonal wavelet transformation, extract respectively low frequency and the high frequency subimage of former Infrared Thermogram; Fig. 2 is two layers of wavelet decomposition schematic diagram of infrared image, and wherein to look like be that LL(LL1 is that one deck decomposes low frequency subgraph picture to low frequency subgraph, and LL2 is two layers and decomposes low frequency subgraph pictures), smoothly approach frame; High frequency imaging comprises: LH vertical information image, and HL horizontal information image, HH diagonal line frame, wherein LH1, HL1, HH1 are that one deck decomposes high frequency subimage, LH2, HL2, HH2 are two layers and decompose high frequency subimage.

Step 3: the low frequency component image that step 1 is decomposed carries out homomorphic filtering, Homomorphic Filtering Algorithm adopts logarithm operation will irradiate component and reflecting component separates, then carries out Fourier transform.Image local details, because luminance component belongs to slow variation, mainly concentrates on low-frequency band relatively, therefore adopts Gauss's high-pass filtering function as homomorphic filtering function, leaches HFS;

Step 4: the high fdrequency component image that step 1 is decomposed carries out fuzzy enhancing, it is evaluation criterion that gradation of image information entropy (formula 9) is take in this fuzzy enhancing, utilizes dynamic self-adapting genetic algorithm (formula 3), (formula 4) to determine the fuzzy parameter F in membership function (formula 6) _dand F _e;

Step 5: the high fdrequency component image that the low frequency component image that homomorphic filtering was processed to step 2 carries out and the fuzzy enhancing of step 3 was processed carries out wavelet reconstruction, this wavelet reconstruction is to carry out orthogonal wavelet inverse transformation with the process contrary with step 1, the Infrared Thermogram being enhanced.

Step 6: according to the running status of the Infrared Thermogram monitoring power equipment of the enhancing obtaining.

The Infrared Thermogram of evaluating the present embodiment for Objective Test On Numberical strengthens effect, adopts contrast, edge strength, information entropy and 4 evaluation indexes of image definition, and it is carried out to objective quantitative test.Table 1 has provided contrast, edge strength, information entropy and the image definition value of the rear image of each algorithm enhancing.

The different algorithm evaluation index contrast tables that strengthen of table 1

Data by table 1 can find out, during the Dynamic Genetics fuzzy enhancement algorithm of application the present embodiment, the contrast of image, information entropy, sharpness obviously increase, and edge strength slightly reduces, and wherein, information entropy increases explanation amount of image information and increases; And strengthen comparing of algorithm with application Genetic-fuzzy, edge strength slightly reduces, this is main because the Dynamic Genetics fuzzy enhancement algorithm of the present embodiment, when image high fdrequency component is carried out to Fuzzy Processing, has carried out suppressing processing for medium-high frequency information relatively low in high fdrequency component.But from total treatment effect, the more original Infrared Thermogram of image that the Dynamic Genetics fuzzy enhancement algorithm of the present embodiment is processed is that relative other enhancing algorithms of contrast, information entropy or other indexes have substantially all been obtained better effect, be more suitable for human eye vision and intuitively respond to resolution, therefore can reach the object that better Infrared Thermogram strengthens.

Vision is directly observed and quantitative analysis results shows, the method of the present embodiment can make the high-frequency information of Infrared Thermogram be enhanced, low-frequency information and noise are inhibited, with respect to additive method, further improve the contrast of target in Infrared Thermogram, kept image self-characteristic, strengthened the visual effect of image, effective dissection to electrical equipment Infrared Thermogram in transformer station is remarkable, is conducive to multi-source image registration, fusion and equipment failure location.

Be noted that above enumerate only for specific embodiments of the invention, obviously the invention is not restricted to above embodiment, have many similar variations thereupon.If all distortion that those skilled in the art directly derives or associates from content disclosed by the invention, all should belong to protection scope of the present invention.

Claims (8)

1. a method of monitoring power equipment running status, it comprises the following steps:

Obtain the Infrared Thermogram of power equipment running status;

Wavelet transformation: described Infrared Thermogram is carried out to wavelet transformation, extract its low frequency component and high fdrequency component;

Homomorphic filtering: described low frequency component is carried out to homomorphic filtering;

Fuzzy enhancing step: described high fdrequency component is carried out to fuzzy enhancing;

Wavelet reconstruction: the high fdrequency component that the low frequency component that homomorphic filtering was processed and fuzzy enhancing were processed is carried out wavelet inverse transformation, the Infrared Thermogram being enhanced;

According to the running status of the Infrared Thermogram monitoring power equipment of the enhancing obtaining.

2. the method for monitoring power equipment running status as claimed in claim 1, is characterized in that, in described fuzzy enhancing step, adopts dynamic self-adapting genetic algorithm to determine the required fuzzy parameter of described fuzzy enhancing.

3. the method for monitoring power equipment running status as claimed in claim 1, is characterized in that, adopts dynamic self-adapting genetic algorithm and take gradation of image information entropy to determine the required fuzzy parameter of described fuzzy enhancing as evaluation criterion.

4. monitor as claimed in claim 2 or claim 3 the method for power equipment running status, it is characterized in that, described dynamic self-adapting genetic algorithm coincidence formula:

$P_c=\left\{\begin{array}{c}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.$

$P_m=\left\{\begin{array}{c}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f)}{f_{\max }-f_{\mathrm{avg}}},f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},f<f_{\mathrm{avg}}\end{array}\right.$

In formula, P _cfor crossover probability; P _mfor variation probability; f _maxmaximum adaptation degree in Wei Meidai colony; f _avgthe average fitness of Wei Meidai colony; F ' is larger fitness in two individualities that will intersect; F is for making a variation individual fitness; P _c1, P _c2, P _m1and P _m2be a constant in interval (0,1).

5. the method for monitoring power equipment running status as claimed in claim 4, is characterized in that P _c1=0.9.

6. the method for monitoring power equipment running status as claimed in claim 4, is characterized in that P _c2=0.6.

7. the method for monitoring power equipment running status as claimed in claim 4, is characterized in that P _m1=0.1.

8. the method for monitoring power equipment running status as claimed in claim 4, is characterized in that P _m2=0.001.