CN116822205B - Rapid fault early warning method for multi-dimensional ring main unit - Google Patents
Rapid fault early warning method for multi-dimensional ring main unit Download PDFInfo
- Publication number
- CN116822205B CN116822205B CN202310777890.6A CN202310777890A CN116822205B CN 116822205 B CN116822205 B CN 116822205B CN 202310777890 A CN202310777890 A CN 202310777890A CN 116822205 B CN116822205 B CN 116822205B
- Authority
- CN
- China
- Prior art keywords
- main unit
- ring main
- overheat
- early warning
- index data
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 40
- 238000007781 pre-processing Methods 0.000 claims abstract description 32
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 42
- 238000004364 calculation method Methods 0.000 claims description 29
- 238000012545 processing Methods 0.000 claims description 28
- 239000011159 matrix material Substances 0.000 claims description 25
- 238000001914 filtration Methods 0.000 claims description 22
- 230000009466 transformation Effects 0.000 claims description 22
- 238000011156 evaluation Methods 0.000 claims description 18
- 230000009467 reduction Effects 0.000 claims description 16
- 230000004927 fusion Effects 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 11
- 230000004913 activation Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 230000007797 corrosion Effects 0.000 claims description 5
- 238000005260 corrosion Methods 0.000 claims description 5
- 238000001514 detection method Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 claims description 5
- 230000007613 environmental effect Effects 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 5
- 230000000717 retained effect Effects 0.000 claims description 3
- 238000011161 development Methods 0.000 abstract description 3
- 238000004891 communication Methods 0.000 description 7
- 230000006399 behavior Effects 0.000 description 6
- 230000006872 improvement Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 238000004378 air conditioning Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000013021 overheating Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of ring main unit fault early warning, and discloses a multi-dimensional ring main unit fault rapid early warning method, which comprises the following steps: collecting ring main unit overheat index data and preprocessing, and inputting the preprocessed ring main unit overheat index data into a ring main unit early warning model to obtain overheat risk early warning values; collecting a continuous partial image of the ring main unit and preprocessing; extracting key parts of the preprocessed ring main unit connection part image; and obtaining a screw loosening result of the ring main unit by using the ring main unit fastening loosening identification model. According to the method, according to the prediction result of the ring main unit overheat index data, the internal temperature value at the next time is compared with the current temperature value, the overheat risk early warning value is dynamically updated, the prediction of overheat fault development trend is realized, multi-scale decomposition is carried out on the image of the key part of screw fastening, and if the decomposition result under the original scale has a larger structural difference with the maximum decomposition result, the screw of the ring main unit is indicated to be loosened.
Description
Technical Field
The invention relates to the technical field of fault early warning, in particular to a method for rapidly early warning faults of a multi-dimensional ring main unit.
Background
The continuous improvement of installed power generation capacity and power grid voltage level puts higher requirements on the safety and reliability of a power system, so that the importance of on-line monitoring and fault diagnosis of power equipment is more and more remarkable. The ring main unit is an important power device in the power distribution network and is widely used for realizing regional closed ring network power supply. However, due to various factors, the ring main unit is extremely easy to overheat in the cabinet, and if the ring main unit is changed into serious faults, the ring main unit can be separated, so that the power supply reliability is affected. The reason for overheat in the cabinet is that the connection part and the screw fastening part are in poor contact, are in a virtual connection state for a long time, have large contact resistance, and generate heat and even burn. Aiming at the problem, the invention provides a multi-dimensional ring main unit fault rapid early warning method for realizing rapid detection and positioning of the safety state of the ring main unit.
Disclosure of Invention
In view of the above, the invention provides a method for fast pre-warning of faults of a multi-dimensional ring main unit, which aims at: 1) Collecting ring main unit overheat index time sequence data of a multidimensional index, and predicting ring main unit overheat index data at the next moment by utilizing a ring main unit early warning model, wherein the collected indexes comprise environmental temperature, contact resistance, flowing current and internal temperature indexes, comparing the predicted internal temperature value at the next moment with the current temperature value according to a prediction result, dynamically updating overheat risk early warning values, comparing the overheat risk early warning values with overheat risk early warning values calculated by other dimension indexes, selecting a larger overheat risk early warning value, and avoiding the problem of inaccurate single-dimension prediction, thereby realizing the prediction of overheat fault development trend; 2) The method comprises the steps of collecting continuous part images of a ring main unit, determining edge pixels, performing expansion corrosion operation on the edge pixels to obtain a plurality of connected domains, calculating the area of each connected domain, deleting the connected domain edge pixels with the connected domain area smaller than a preset area threshold value, removing redundant edge information, further performing affine transformation processing on the coordinates of the connected domain edge pixels by combining with a standard ring main unit image, wherein the farther the coordinate distance between the coordinates before and after affine transformation is, the lower the matching degree between the current coordinates and the standard ring main unit coordinate is, obtaining inner points with higher matching degree and screw fastening key part images by combining with an evaluation value screening result, performing multi-scale decomposition on the screw fastening key part images, and performing screw loosening early warning if a large structural difference exists between the decomposition result under the original scale 1 and the maximum decomposition result, and performing screw loosening early warning on the ring main unit, thereby realizing ring main unit overheating early warning and screw loosening early warning.
In order to achieve the above purpose, the invention provides a method for rapidly pre-warning faults of a multi-dimensional ring main unit, which comprises the following steps:
s1: constructing a ring main unit overheat index system, collecting ring main unit overheat index data and preprocessing the ring main unit overheat index data to obtain preprocessed ring main unit overheat index data;
s2: constructing a ring main unit early warning model, and inputting preprocessed ring main unit overheat index data into the model to obtain overheat risk early warning values;
s3: if the overheat risk early warning value exceeds a preset threshold value, collecting the image of the continuing part of the ring main unit and preprocessing the image to obtain the preprocessed image of the continuing part of the ring main unit;
s4: extracting key parts of the preprocessed ring main unit connection part image to obtain a screw fastening key part image;
s5: and constructing a ring main unit fastening loosening identification model, wherein the constructed model takes an image of a key part of screw fastening as input and a screw loosening result as output, the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and if screw loosening exists, screw loosening early warning is carried out.
As a further improvement of the present invention:
optionally, in the step S1, a ring main unit overheat index system is constructed, and ring main unit overheat index data is collected, including:
Building a ring main unit overheat index system, wherein the ring main unit overheat index system is as follows:
(w 0 ,R,I,w)
wherein:
w 0 the environmental temperature index of the ring main unit is represented;
w represents an internal temperature index of the ring main unit;
r represents a contact resistance index of a bus connection point of the ring main unit;
i represents the index of the flowing current of the ring main unit;
the method comprises the steps of collecting ring main unit overheat index data based on a ring main unit overheat index system, wherein the collected ring main unit overheat index data are as follows:
{x(t)＝(x(t 0 ),x(t 1 ),...,x(t s ),...,x(t S ))|x∈{w 0 ,R,I,w}}
wherein:
x (t) represents time series data of the collected index x, t 0 Indicating the current time, t 1 ～t S The historical time is represented, and the time interval between adjacent times is Δt.
Optionally, preprocessing the ring main unit overheat index data in the step S1 includes:
preprocessing the ring main unit overheat index data, wherein the preprocessing flow is as follows:
s11: carrying out normalization processing on the acquired data of the index x at any moment, wherein the normalization processing formula is as follows:
wherein:
x(t s ) Representing t s Acquisition data of time index x, x ′ (t s ) Represents x (t) s ) Is a normalization processing result of (a);
max x represents the maximum value, min, of the set index x x Representing the minimum value of the set index x;
s12: converting the normalized ring main unit overheat index data into a matrix form X:
Wherein:
w 0 ′ (t s ),R ′ (t s ),I ′ (t s ),w ′ (t S ) Respectively t after normalization processing S Moment ring main unit overheat index w 0 Collecting data of R, I and w;
and taking X as a preprocessing result of the ring main unit overheat index data.
Optionally, in the step S2, the preprocessed ring main unit overheat index data is input into a ring main unit early warning model to obtain an overheat risk early warning value, which includes:
the method comprises the steps of constructing a ring main unit early warning model, wherein the ring main unit early warning model comprises an input layer, a characteristic calculation layer and a characteristic fusion layer, the input layer is used for receiving preprocessed ring main unit overheat index data, the characteristic calculation layer comprises three convolution layers and is used for carrying out convolution calculation on the preprocessed ring main unit overheat index data, the characteristic fusion layer is used for fusing calculation results of the three convolution layers to obtain ring main unit overheat index data at a prediction moment, and overheat risk early warning values are obtained based on ring main unit overheat index data at the prediction moment;
the preprocessed ring main unit overheat index data is input into a ring main unit early warning model to obtain overheat risk early warning values, and the overheat risk early warning value calculation flow based on the ring main unit early warning model is as follows:
the input layer of the ring main unit early warning model receives the preprocessed ring main unit overheat index data X;
Setting the current iteration number as K, the initial value of K as 1, and the maximum iteration number as K, wherein the result obtained by the kth iteration is C k The result obtained by each iteration comprises the data of four ring main unit overheat indexes, namely ring main unit overheat index data at the predicted moment;
the iteration formula of the kth iteration is as follows:
C k ＝σ 2 ([X,C k-1 ]w 4 +b 4 )tanh(F 3,k-1 )
F 3,k-1 ＝σ 2 [(F 1,k-1 +F 2,k-1 )w 3 +b 3 ]
F 1,k-1 ＝σ 1 ([X,C k-1 ]w 1 +b 1 )
F 2,k-1 ＝σ 1 ([X,C k-1 ]w 2 +b 2 )
wherein:
w 1 ,w 2 ,w 3 representing a characteristic meterCalculating weight parameters of three convolution layers in the layer, b 1 ,b 2 ,b 3 Representing bias parameters of three convolution layers in the feature calculation layer;
w 4 weight parameters representing feature fusion layer, b 4 Representing bias parameters of the feature fusion layer;
σ 1 (·),σ 2 (. Cndot.) represents an activation function; in the embodiment of the invention, sigma 1 (. Cndot.) is a softmax activation function, σ 2 (. Cndot.) is a ReLU activation function;
C 0 the first column in X is the pre-processed ring main unit overheat index data at the current moment;
let k=k+1, repeat the iterative formula of the kth iteration until the iterative result of the kth iteration is obtained, namely the ring main unit overheat index data (w 0 * ,R * ,I * ,w * ) Wherein w is 0 * Ambient temperature index data representing the predicted time, R * Contact resistance index data representing ring main unit bus connection point at prediction time, I * Ring main unit flowing current index data w representing prediction time * The ring main unit internal temperature index data representing the prediction time;
ring main unit overheat index data (w) based on prediction time 0 * ,R * ,I * ,w * ) And calculating to obtain an overheat risk early warning value:
wherein:
and representing the overheat risk early warning value.
Optionally, if the overheat risk early warning value exceeds the preset threshold in the step S3, collecting the image of the continuing part of the ring main unit and performing preprocessing, including:
if the overheat risk early warning value exceeds a preset threshold value, collecting and preprocessing a ring main unit connection part image, wherein the ring main unit connection part image is an image of a ring main unit screw connection position, and the preprocessing flow of the ring main unit connection part image is as follows:
s31: carrying out graying treatment on any pixel I (I, j) of the image of the ring main unit connection part, wherein the gray value of the pixel I (I, j) is as follows:
g(i,j)＝max{I R (i,j),I G (i,j),I B (i,j)}
wherein:
I R (i,j),I G (i,j),I B (I, j) are the color values of pixel I (I, j) in the RGB color channel, respectively;
g (I, j) represents the gray value of pixel I (I, j), and pixel I (I, j) represents the pixel of the ith row and the jth column in the image of the successive part of the ring main unit;
s32: filtering and noise reduction processing is carried out on the image of the ring main unit connection part obtained by the graying processing, and the filtering and noise reduction processing flow is as follows:
a gaussian kernel of 3×3 in size and 1 in standard deviation was set in the form of:
For any pixel in the image of the continuous part of the ring main unit after the graying treatment, multiplying a 3 multiplied by 3 pixel area taking the pixel as the center by a Gaussian kernel, and adding all numbers in the product result to obtain a filtering treatment result of the pixel, namely a gray value after the filtering treatment;
s33: for a 3×3 gray matrix a centered on an arbitrary pixel I (I, j), the Sobel operator S is used 1 And S is 2 Computing gradient matrix g for pixel I (I, j) 12 (i,j)：
Wherein:
* Representing a convolution process;
g 1 (I, j) represents a gradient matrix of the pixel I (I, j) in the horizontal direction;
g 2 (I, j) represents a gradient matrix of the pixel I (I, j) in the vertical direction;
s34: calculate g up (i, j) and g down (i, j) if g 12 The sum of the elements in (i, j) is greater than g up (i, j) and g down The sum of the elements of (I, j) then pixel I (I, j) is an edge pixel and the gray value of pixel I (I, j) is retained, otherwise it is set to 0, g up (i, j) and g down The calculation formula of (i, j) is:
wherein:
i represent L1 norm;
edge pixel detection is carried out on all pixels, so that an edge diagram of a connection part of the ring main unit is obtained;
s35: performing expansion corrosion operation on edge pixels in the edge map to obtain a plurality of connected domains, and calculating the area of each connected domain;
s36: deleting connected domain edge pixels with the connected domain area smaller than a preset area threshold value, removing redundant edge information, marking the connected domain edge pixel positions of the reserved connected domain in the image of the connected part of the ring main unit after filtering and noise reduction treatment, and taking the image of the connected part of the ring main unit after filtering and noise reduction treatment, which marks the connected domain edge pixel positions, as the image of the connected part of the ring main unit after pretreatment.
Optionally, in the step S4, extracting a key part of the preprocessed image of the connection part of the ring main unit includes:
extracting key parts of the preprocessed ring main unit connection part image, wherein the key part extraction process comprises the following steps:
s41: calculating gradient directions of connected domain edge pixels in the preprocessed ring main unit continuous part images, wherein the connected domain edge pixels I ′ (i * ,j * ) Gradient direction θ (i) * ,j * ) The calculation formula is as follows:
wherein:
g ′ (i * ,j * ) Representing connected domain edge pixels I ′ (i * ,j * ) Is used for filtering the gray value after noise reduction;
s42: for connected domain edge pixel I ′ (i * ,j * ) The coordinates of (i) are rotated to obtain rotated coordinates ((i) * ) ′ ,(j * ) ′ )：
S43: affine transformation is performed on the rotated connected domain edge pixel coordinates, wherein the coordinates ((i) * ) ′ ,(j * ) ′ ) The affine transformation formula of (2) is:
wherein:
(i ′ ,j ′ ) For the coordinates ((i) * ) ′ ,(j * ) ′ ) Affine transformation results of (a);
h is an affine transformation matrix; in the embodiment of the invention, the invention collects the connected domain of a plurality of groups of sample imagesEdge pixel coordinates and connected domain edge pixel coordinates of standard ring main unit image form matrixes B and B ′ Wherein the matrix is divided into three rows, the coordinates of the first behavior in the horizontal direction, the coordinates of the second behavior in the vertical direction and the third behavior 1, the calculation formula of the affine transformation matrix is H=B ′ B T (B ′ B T ) -1 ；
S44: calculating Euclidean distance between each connected domain edge pixel coordinate and affine transformed coordinate, wherein the Euclidean distance between the D-th connected domain edge pixel coordinate and affine transformed coordinate is dis (D), the total number of the connected domain edge pixel coordinates is D, and the self-adaptive threshold is initialized to delta=0.1;
s45: marking the edge pixel coordinates of the connected domain with the Euclidean distance smaller than the self-adaptive threshold value as an inner point, otherwise marking as an outer point, and calculating the current evaluation value:
wherein:
n in represents the number of interior points, n out Representing the number of outliers;
Ω in representing the set of interior points, Ω out Represents the outlier set, dis (d) in ) Representing the inner point d in Euclidean distance from affine transformed coordinates, dis (d out ) Representing the outer point d out Euclidean distance from affine transformed coordinates;
representing the average Euclidean distance between the edge pixel coordinates of the D connected domains and the affine transformed coordinates;
storing the current evaluation value and the interior point set;
s46: let δ=δ+0.3, return to step S45 until δ=3;
s47: traversing the currently stored evaluation value and the corresponding interior point set, and selecting the interior point set with the minimum evaluation value; and regarding a plurality of connected domain images in the preprocessed ring main unit continuous part images, if the internal point proportion of the edges of the connected domain images reaches 30%, taking the connected domain images as screw fastening key part images.
Optionally, in the step S5, the ring main unit tightening loosening identification model is used to obtain a screw loosening result of the ring main unit, including:
constructing a ring main unit fastening loosening identification model, wherein the constructed model takes a screw fastening key part image as input and a screw loosening result as output, and the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and the screw loosening result identification flow based on the ring main unit fastening loosening identification model is as follows:
s51: calculating to obtain the gradient direction of each pixel in the screw fastening key part image;
s52: and decomposing the image of the key part of screw fastening, wherein the decomposing formula is as follows:
wherein:
b represents a decomposition scale, C (B) represents a decomposition result under the scale B, wherein B E [1, B ], and B represents a maximum decomposition scale;
g ′ (h, y) represents the pixel I in the screw tightening key part image ′ The filtered and denoised gray value of (h, y);
θ (h, y) represents the pixel I ′ A gradient direction of (h, y);
s53: calculating to obtain maximum decomposition results C of B decomposition scales max And minimum decomposition result C min ；
S54: if it isIf the difference is smaller than the preset threshold value, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is not larger, the screw of the ring main unit is not loosened, otherwise, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is larger, and the screw of the ring main unit is stored And when the screw is loosened, the screw loosening early warning is carried out if the screw is loosened.
In order to solve the above-described problems, the present invention provides an electronic apparatus including:
a memory storing at least one instruction;
the communication interface is used for realizing the communication of the electronic equipment; a kind of electronic device with high-pressure air-conditioning system
And the processor executes the instructions stored in the memory to realize the multi-dimensional ring main unit fault rapid early warning method.
In order to solve the above problems, the present invention further provides a computer readable storage medium, where at least one instruction is stored, where the at least one instruction is executed by a processor in an electronic device to implement the foregoing method for fast early warning of a failure of a multi-dimensional ring main unit.
Compared with the prior art, the invention provides a rapid fault early warning method for a multi-dimensional ring main unit, which has the following advantages:
firstly, the scheme provides a ring main unit overheat early warning method, wherein preprocessed ring main unit overheat index data are input into a ring main unit early warning model to obtain overheat risk early warning values, and the overheat risk early warning value calculation flow based on the ring main unit early warning model is as follows: the input layer of the ring main unit early warning model receives the preprocessed ring main unit overheat index data X; setting the current iteration number as K, the initial value of K as 1, and the maximum iteration number as K, wherein the result obtained by the kth iteration is C k The result obtained by each iteration comprises the data of four ring main unit overheat indexes, namely ring main unit overheat index data at the predicted moment; the iteration formula of the kth iteration is as follows:
C k ＝σ 2 ([X,C k-1 ]w 4 +b 4 )tanh(F 3,k-1 )
F 3,k-1 ＝σ 2 [(F 1,k-1 +F 2,k-1 )w 3 +b 3 ]
F 1,k-1 ＝σ 1 ([X,C k-1 ]w 1 +b 1 )
F 2,k-1 ＝σ 1 ([X,C k-1 ]w 2 +b 2 )
wherein: w (w) 1 ,w 2 ,w 3 Weight parameters representing three convolutional layers in the feature computation layer, b 1 ,b 2 ,b 3 Representing bias parameters of three convolution layers in the feature calculation layer; w (w) 4 Weight parameters representing feature fusion layer, b 4 Representing bias parameters of the feature fusion layer;
σ 1 (·),σ 2 (. Cndot.) represents an activation function; c (C) 0 The first column in X is the pre-processed ring main unit overheat index data at the current moment; let k=k+1, repeat the iterative formula of the kth iteration until the iterative result of the kth iteration is obtained, namely the ring main unit overheat index data (w 0 * ,R * ,I * ,w * ) Wherein w is 0 * Ambient temperature index data representing the predicted time, R * Contact resistance index data representing ring main unit bus connection point at prediction time, I * Ring main unit flowing current index data w representing prediction time * The ring main unit internal temperature index data representing the prediction time; ring main unit overheat index data (w) based on prediction time 0 * ,R * ,I * ,w * ) And calculating to obtain an overheat risk early warning value:
wherein:and representing the overheat risk early warning value. According to the scheme, ring main unit overheat index time sequence data of multidimensional indexes are collected, ring main unit overheat index data at the next moment are predicted by using a ring main unit early warning model, wherein the collected indexes comprise environmental temperature, contact resistance, flowing current and internal temperature indexes, the internal temperature value at the next moment is predicted to be compared with the current temperature value according to a prediction result, and an overheat risk early warning value is dynamically updated, and the method is the same as the method And comparing overheat risk early-warning values calculated by other dimension indexes, selecting a larger overheat risk early-warning value, and avoiding the problem of inaccurate single-dimension prediction, thereby realizing the prediction of overheat fault development trend.
Meanwhile, the scheme provides a ring main unit screw loosening early warning, and key part extraction is carried out on preprocessed ring main unit continuous part images, wherein the key part extraction flow is as follows: calculating gradient directions of connected domain edge pixels in the preprocessed ring main unit continuous part images, wherein the connected domain edge pixels I ′ (i * ,j * ) Gradient direction θ (i) * ,j * ) The calculation formula is as follows:
wherein: g ′ (i * ,j * ) Representing connected domain edge pixels I ′ (i * ,j * ) Is used for filtering the gray value after noise reduction; for connected domain edge pixel I ′ (i * ,j * ) The coordinates of (i) are rotated to obtain rotated coordinates ((i) * ) ′ ,(j * ) ′ )：
Affine transformation is performed on the rotated connected domain edge pixel coordinates, wherein the coordinates ((i) * ) ′ ,(j * ) ′ ) The affine transformation formula of (2) is
Wherein: (i) ′ ,j ′ ) For the coordinates ((i) * ) ′ ,(j * ) ′ ) Affine transformation results of (a); h is an affine transformation matrix; calculating the Euclidean distance between each connected domain edge pixel coordinate and affine transformed coordinate, wherein the d connected domain edge pixel coordinateThe Euclidean distance between the pixel coordinates and the affine transformed coordinates is dis (D), the total number of the pixel coordinates of the edge of the connected domain is D, and the self-adaptive threshold value is initialized to delta=0.1; marking the edge pixel coordinates of the connected domain with the Euclidean distance smaller than the self-adaptive threshold value as an inner point, otherwise marking as an outer point, and calculating the current evaluation value:
Wherein: n is n in Represents the number of interior points, n out Representing the number of outliers; omega shape in Representing the set of interior points, Ω out Represents the outlier set, dis (d) in ) Representing the inner point d in Euclidean distance from affine transformed coordinates, dis (d out ) Representing the outer point d out Euclidean distance from affine transformed coordinates;representing the average Euclidean distance between the edge pixel coordinates of the D connected domains and the affine transformed coordinates; storing the current evaluation value and the interior point set; let δ=δ+0.3; traversing the currently stored evaluation value and the corresponding interior point set, and selecting the interior point set with the minimum evaluation value; and regarding a plurality of connected domain images in the preprocessed ring main unit continuous part images, if the internal point proportion of the edges of the connected domain images reaches 30%, taking the connected domain images as screw fastening key part images. According to the scheme, through collecting continuous part images of a ring main unit, determining edge pixels, performing expansion corrosion operation on the edge pixels to obtain a plurality of connected domains, calculating the area of each connected domain, deleting the connected domain edge pixels with the connected domain area smaller than a preset area threshold value, removing redundant edge information, further performing affine transformation processing combining standard ring main unit images on the coordinates of the connected domain edge pixels, wherein the far the coordinate distance between the front and the back of affine transformation is, the lower the matching degree between the current coordinates and the standard ring main unit coordinates is, the inner points with higher matching degree and screw fastening key part images are obtained by combining evaluation value screening results, multi-scale decomposition is performed on the screw fastening key part images, and if the decomposition junction is under the original scale 1, the decomposition junction is formed The result has larger structural difference with the maximum decomposition result, the screw of the ring main unit is loosened, and the screw loosening early warning is carried out, so that the ring main unit and the screw loosening early warning are realized.
Drawings
Fig. 1 is a schematic flow chart of a method for fast pre-warning of a multi-dimensional ring main unit fault according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an electronic device for implementing a method for fast early warning of a fault of a multi-dimensional ring main unit according to an embodiment of the present invention.
The achievement of the objects, functional features and advantages of the present invention will be further described with reference to the accompanying drawings, in conjunction with the embodiments.
Detailed Description
It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The embodiment of the application provides a rapid fault early warning method for a multi-dimensional ring main unit. The execution main body of the multi-dimensional ring main unit fault quick warning method comprises, but is not limited to, at least one of a server side, a terminal and the like which can be configured to execute the electronic equipment of the method provided by the embodiment of the application. In other words, the method for quickly warning the fault of the multi-dimensional ring main unit can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The service end includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
Example 1:
s1: and constructing a ring main unit overheat index system, collecting ring main unit overheat index data and preprocessing the ring main unit overheat index data to obtain preprocessed ring main unit overheat index data.
The step S1 is to construct a ring main unit overheat index system, collect ring main unit overheat index data, and comprises:
building a ring main unit overheat index system, wherein the ring main unit overheat index system is as follows:
(w 0 ,R,I,w)
wherein:
w 0 the environmental temperature index of the ring main unit is represented;
w represents an internal temperature index of the ring main unit;
r represents a contact resistance index of a bus connection point of the ring main unit;
i represents the index of the flowing current of the ring main unit;
the method comprises the steps of collecting ring main unit overheat index data based on a ring main unit overheat index system, wherein the collected ring main unit overheat index data are as follows:
{x(t)＝(x(t 0 ),x(t 1 ),...,x(t s ),...,x(t S ))|x∈{w 0 ,R,I,w}}
wherein:
x (t) represents time series data of the collected index x, t 0 Indicating the current time, t 1 ～t S The historical time is represented, and the time interval between adjacent times is Δt.
In the step S1, preprocessing is performed on the ring main unit overheat index data, and the method comprises the following steps:
preprocessing the ring main unit overheat index data, wherein the preprocessing flow is as follows:
s11: carrying out normalization processing on the acquired data of the index x at any moment, wherein the normalization processing formula is as follows:
Wherein:
x(t s ) Representing t s Acquisition data of time index x, x ′ (t s ) Represents x (t) s ) Is a normalization processing result of (a);
max x represents the maximum value, min, of the set index x x Representing the minimum value of the set index x;
s12: converting the normalized ring main unit overheat index data into a matrix form X:
wherein:
w 0 ′ (t S ),R ′ (t S ),I ′ (t S ),w ′ (t S ) Respectively t after normalization processing S Moment ring main unit overheat index w 0 Collecting data of R, I and w;
and taking X as a preprocessing result of the ring main unit overheat index data.
S2: and constructing a ring main unit early warning model, and inputting the preprocessed ring main unit overheat index data into the model to obtain overheat risk early warning values.
In the step S2, the preprocessed ring main unit overheat index data is input into a ring main unit early warning model to obtain overheat risk early warning values, and the method comprises the following steps:
the method comprises the steps of constructing a ring main unit early warning model, wherein the ring main unit early warning model comprises an input layer, a characteristic calculation layer and a characteristic fusion layer, the input layer is used for receiving preprocessed ring main unit overheat index data, the characteristic calculation layer comprises three convolution layers and is used for carrying out convolution calculation on the preprocessed ring main unit overheat index data, the characteristic fusion layer is used for fusing calculation results of the three convolution layers to obtain ring main unit overheat index data at a prediction moment, and overheat risk early warning values are obtained based on ring main unit overheat index data at the prediction moment;
The preprocessed ring main unit overheat index data is input into a ring main unit early warning model to obtain overheat risk early warning values, and the overheat risk early warning value calculation flow based on the ring main unit early warning model is as follows:
the input layer of the ring main unit early warning model receives the preprocessed ring main unit overheat index data X;
setting the current iteration number as K, the initial value of K as 1, and the maximum iteration number as K, wherein the result obtained by the kth iteration is C k The result obtained by each iteration comprises the data of four ring main unit overheat indexes, namely ring main unit overheat index data at the predicted moment;
the iteration formula of the kth iteration is as follows:
C k ＝σ 2 ([X,C k-1 ]w 4 +b 4 )tanh(F 3,k-1 )
F 3,k-1 ＝σ 2 [(F 1,k-1 +F 2,k-1 )w 3 +b 3 ]
F 1,k-1 ＝σ 1 ([X,C k-1 ]w 1 +b 1 )
F 2,k-1 ＝σ 1 ([X,C k-1 ]w 2 +b 2 )
wherein:
w 1 ,w 2 ,w 3 weight parameters representing three convolutional layers in the feature computation layer, b 1 ,b 2 ,b 3 Representing bias parameters of three convolution layers in the feature calculation layer;
w 4 weight parameters representing feature fusion layer, b 4 Representing bias parameters of the feature fusion layer;
σ 1 (·),σ 2 (. Cndot.) represents an activation function; in the embodiment of the invention, sigma 1 (. Cndot.) is a softmax activation function, σ 2 (. Cndot.) is a ReLU activation function;
C 0 the first column in X is the pre-processed ring main unit overheat index data at the current moment;
let k=k+1, repeat the iterative formula of the kth iteration until the iterative result of the kth iteration is obtained, namely the ring main unit overheat index data (w 0 * ,R * ,I * ,w * ) Wherein w is 0 * Ambient temperature index data representing the predicted time,
R * contact resistance index data representing ring main unit bus connection point at prediction time, I * Ring main unit flowing current index data w representing prediction time * The ring main unit internal temperature index data representing the prediction time;
ring main unit overheat index data (w) based on prediction time 0 * ,R * ,I * ,w * ) And calculating to obtain an overheat risk early warning value:
wherein:
and representing the overheat risk early warning value.
S3: if the overheat risk early warning value exceeds a preset threshold value, collecting and preprocessing the image of the ring main unit connection part to obtain the preprocessed image of the ring main unit connection part.
And step S3, if the overheat risk early warning value exceeds a preset threshold value, collecting the image of the continuing part of the ring main unit and preprocessing the image, wherein the step comprises the following steps:
if the overheat risk early warning value exceeds a preset threshold value, collecting and preprocessing a ring main unit connection part image, wherein the ring main unit connection part image is an image of a ring main unit screw connection position, and the preprocessing flow of the ring main unit connection part image is as follows:
s31: carrying out graying treatment on any pixel I (I, j) of the image of the ring main unit connection part, wherein the gray value of the pixel I (I, j) is as follows:
g(i,j)＝max{I R (i,j),I G (i,j),I B (i,j)}
Wherein:
I R (i,j),I G (i,j),I B (I, j) are the color values of pixel I (I, j) in the RGB color channel, respectively;
g (I, j) represents the gray value of pixel I (I, j), and pixel I (I, j) represents the pixel of the ith row and the jth column in the image of the successive part of the ring main unit;
s32: filtering and noise reduction processing is carried out on the image of the ring main unit connection part obtained by the graying processing, and the filtering and noise reduction processing flow is as follows:
a gaussian kernel of 3×3 in size and 1 in standard deviation was set in the form of:
for any pixel in the image of the continuous part of the ring main unit after the graying treatment, multiplying a 3 multiplied by 3 pixel area taking the pixel as the center by a Gaussian kernel, and adding all numbers in the product result to obtain a filtering treatment result of the pixel, namely a gray value after the filtering treatment;
s33: for a 3×3 gray matrix a centered on an arbitrary pixel I (I, j), the Sobel operator S is used 1 And S is 2 Computing gradient matrix g for pixel I (I, j) 12 (i,j)：
Wherein:
* Representing a convolution process;
g 1 (I, j) represents a gradient matrix of the pixel I (I, j) in the horizontal direction;
g 2 (I, j) represents a gradient matrix of the pixel I (I, j) in the vertical direction;
s34: calculate g up (i, j) and g down (i, j) if g 12 The sum of the elements in (i, j) is greater than g up (i, j) and g down The sum of the elements of (I, j) then pixel I (I, j) is an edge pixel and the gray value of pixel I (I, j) is retained, otherwise it is set to 0, g up (i, j) and g down The calculation formula of (i, j) is:
wherein:
i represent L1 norm;
edge pixel detection is carried out on all pixels, so that an edge diagram of a connection part of the ring main unit is obtained;
s35: performing expansion corrosion operation on edge pixels in the edge map to obtain a plurality of connected domains, and calculating the area of each connected domain;
s36: deleting connected domain edge pixels with the connected domain area smaller than a preset area threshold value, removing redundant edge information, marking the connected domain edge pixel positions of the reserved connected domain in the image of the connected part of the ring main unit after filtering and noise reduction treatment, and taking the image of the connected part of the ring main unit after filtering and noise reduction treatment, which marks the connected domain edge pixel positions, as the image of the connected part of the ring main unit after pretreatment.
S4: and extracting key parts of the preprocessed ring main unit connection part image to obtain a screw fastening key part image.
And in the step S4, extracting key parts of the preprocessed image of the continuous part of the ring main unit, wherein the method comprises the following steps:
extracting key parts of the preprocessed ring main unit connection part image, wherein the key part extraction process comprises the following steps:
s41: calculating gradient directions of connected domain edge pixels in the preprocessed ring main unit continuous part images, wherein the connected domain edge pixels I ′ (i * ,j * ) Gradient direction θ (i) * ,j * ) The calculation formula is as follows:
wherein:
g ′ (i * ,j * ) Representing connected domain edge pixels I ′ (i * ,j * ) Is used for filtering the gray value after noise reduction;
s42: for connected domain edge pixel I ′ (i * ,j * ) The coordinates of (i) are rotated to obtain rotated coordinates ((i) * ) ′ ,(j * ) ′ )：
S43: affine transformation is performed on the rotated connected domain edge pixel coordinates, wherein the coordinates ((i) * ) ′ ,(j * ) ′ ) The affine transformation formula of (2) is:
wherein:
(i ′ ,j ′ ) For the coordinates ((i) * ) ′ ,(j * ) ′ ) Affine transformation results of (a);
h is an affine transformation matrix; in the embodiment of the invention, the matrix B and the matrix B are formed by collecting the pixel coordinates of the edges of the connected domains of a plurality of groups of sample images and the pixel coordinates of the edges of the connected domains of the standard ring main unit images ′ Wherein the matrix is divided into three rows, the coordinates of the first behavior in the horizontal direction, the coordinates of the second behavior in the vertical direction and the third behavior 1, the calculation formula of the affine transformation matrix is H=B ′ B T (B ′ B T ) -1 ；
S44: calculating Euclidean distance between each connected domain edge pixel coordinate and affine transformed coordinate, wherein the Euclidean distance between the D-th connected domain edge pixel coordinate and affine transformed coordinate is dis (D), the total number of the connected domain edge pixel coordinates is D, and the self-adaptive threshold is initialized to delta=0.1;
s45: marking the edge pixel coordinates of the connected domain with the Euclidean distance smaller than the self-adaptive threshold value as an inner point, otherwise marking as an outer point, and calculating the current evaluation value:
/>
Wherein:
n in represents the number of interior points, n out Representing the number of outliers;
Ω in representing the set of interior points, Ω out Represents the outlier set, dis (d) in ) Representing the inner point d in Euclidean distance from affine transformed coordinates, dis (d out ) Representing the outer point d out Euclidean distance from affine transformed coordinates;
representing the average Euclidean distance between the edge pixel coordinates of the D connected domains and the affine transformed coordinates;
storing the current evaluation value and the interior point set;
s46: let δ=δ+0.3, return to step S45 until δ=3;
s47: traversing the currently stored evaluation value and the corresponding interior point set, and selecting the interior point set with the minimum evaluation value; and regarding a plurality of connected domain images in the preprocessed ring main unit continuous part images, if the internal point proportion of the edges of the connected domain images reaches 30%, taking the connected domain images as screw fastening key part images.
S5: and constructing a ring main unit fastening loosening identification model, wherein the constructed model takes an image of a key part of screw fastening as input and a screw loosening result as output, the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and if screw loosening exists, screw loosening early warning is carried out.
In the step S5, a ring main unit fastening loosening identification model is utilized to obtain a screw loosening result of the ring main unit, and the method comprises the following steps:
Constructing a ring main unit fastening loosening identification model, wherein the constructed model takes a screw fastening key part image as input and a screw loosening result as output, and the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and the screw loosening result identification flow based on the ring main unit fastening loosening identification model is as follows:
s51: calculating to obtain the gradient direction of each pixel in the screw fastening key part image;
s52: and decomposing the image of the key part of screw fastening, wherein the decomposing formula is as follows:
wherein:
b represents a decomposition scale, C (B) represents a decomposition result under the scale B, wherein B E [1, B ], and B represents a maximum decomposition scale;
g ′ (h, y) represents the pixel I in the screw tightening key part image ′ The filtered and denoised gray value of (h, y);
θ (h, y) represents the pixel I ′ A gradient direction of (h, y);
s53: calculating to obtain maximum decomposition results C of B decomposition scales max And minimum decomposition result C min ；
S54: if it isIf the difference is smaller than the preset threshold value, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is not larger, the screw of the ring main unit is not loosened, otherwise, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is larger, the screw of the ring main unit is loosened, and if the screw is loosened, screw loosening early warning is carried out.
Example 2:
fig. 2 is a schematic structural diagram of an electronic device for implementing a method for fast early warning of a fault of a multi-dimensional ring main unit according to an embodiment of the present invention.
The electronic device 1 may comprise a processor 10, a memory 11, a communication interface 13 and a bus, and may further comprise a computer program, such as program 12, stored in the memory 11 and executable on the processor 10.
The memory 11 includes at least one type of readable storage medium, including flash memory, a mobile hard disk, a multimedia card, a card memory (e.g., SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device 1, such as a removable hard disk of the electronic device 1. The memory 11 may in other embodiments also be an external storage device of the electronic device 1, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the electronic device 1. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device 1. The memory 11 may be used not only for storing application software installed in the electronic device 1 and various types of data, such as codes of the program 12, but also for temporarily storing data that has been output or is to be output.
The processor 10 may be comprised of integrated circuits in some embodiments, for example, a single packaged integrated circuit, or may be comprised of multiple integrated circuits packaged with the same or different functions, including one or more central processing units (Central Processing unit, CPU), microprocessors, digital processing chips, graphics processors, combinations of various control chips, and the like. The processor 10 is a Control Unit (Control Unit) of the electronic device, connects various components of the entire electronic device using various interfaces and lines, executes or executes programs or modules (a program 12 for implementing fast warning of a failure of the multi-dimensional ring main Unit, etc.) stored in the memory 11, and invokes data stored in the memory 11 to perform various functions of the electronic device 1 and process the data.
The communication interface 13 may comprise a wired interface and/or a wireless interface (e.g. WI-FI interface, bluetooth interface, etc.), typically used to establish a communication connection between the electronic device 1 and other electronic devices and to enable connection communication between internal components of the electronic device.
The bus may be a peripheral component interconnect standard (peripheral component interconnect, PCI) bus or an extended industry standard architecture (extended industry standard architecture, EISA) bus, among others. The bus may be classified as an address bus, a data bus, a control bus, etc. The bus is arranged to enable a connection communication between the memory 11 and at least one processor 10 etc.
Fig. 2 shows only an electronic device with components, it being understood by a person skilled in the art that the structure shown in fig. 2 does not constitute a limitation of the electronic device 1, and may comprise fewer or more components than shown, or may combine certain components, or may be arranged in different components.
For example, although not shown, the electronic device 1 may further include a power source (such as a battery) for supplying power to each component, and preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that functions of charge management, discharge management, power consumption management, and the like are implemented through the power management device. The power supply may also include one or more of any of a direct current or alternating current power supply, recharging device, power failure detection circuit, power converter or inverter, power status indicator, etc. The electronic device 1 may further include various sensors, bluetooth modules, wi-Fi modules, etc., which will not be described herein.
The electronic device 1 may optionally further comprise a user interface, which may be a Display, an input unit, such as a Keyboard (Keyboard), or a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like. The display may also be referred to as a display screen or display unit, as appropriate, for displaying information processed in the electronic device 1 and for displaying a visual user interface.
It should be understood that the embodiments described are for illustrative purposes only and are not limited to this configuration in the scope of the patent application.
The program 12 stored in the memory 11 of the electronic device 1 is a combination of instructions that, when executed in the processor 10, may implement:
constructing a ring main unit overheat index system, collecting ring main unit overheat index data and preprocessing the ring main unit overheat index data to obtain preprocessed ring main unit overheat index data;
constructing a ring main unit early warning model, and inputting preprocessed ring main unit overheat index data into the model to obtain overheat risk early warning values;
if the overheat risk early warning value exceeds a preset threshold value, collecting the image of the continuing part of the ring main unit and preprocessing the image to obtain the preprocessed image of the continuing part of the ring main unit;
extracting key parts of the preprocessed ring main unit connection part image to obtain a screw fastening key part image;
and constructing a ring main unit fastening loosening identification model, wherein the constructed model takes an image of a key part of screw fastening as input and a screw loosening result as output, the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and if screw loosening exists, screw loosening early warning is carried out.
Specifically, the specific implementation method of the above instruction by the processor 10 may refer to descriptions of related steps in the corresponding embodiments of fig. 1 to 2, which are not repeated herein.
It should be noted that, the foregoing reference numerals of the embodiments of the present invention are merely for describing the embodiments, and do not represent the advantages and disadvantages of the embodiments. And the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, apparatus, article or method that comprises the element.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) as described above, comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The foregoing description is only of the preferred embodiments of the present invention, and is not intended to limit the scope of the invention, but rather is intended to cover any equivalents of the structures or equivalent processes disclosed herein or in the alternative, which may be employed directly or indirectly in other related arts.
Claims (6)
1. A multi-dimensional ring main unit fault rapid early warning method is characterized by comprising the following steps:
s1: constructing a ring main unit overheat index system, collecting ring main unit overheat index data and preprocessing the ring main unit overheat index data to obtain preprocessed ring main unit overheat index data;
s2: constructing a ring main unit early warning model, and inputting preprocessed ring main unit overheat index data into the model to obtain overheat risk early warning values;
s3: if the overheat risk early warning value exceeds a preset threshold value, collecting the image of the continuing part of the ring main unit and preprocessing the image to obtain the preprocessed image of the continuing part of the ring main unit;
s4: extracting key parts of the preprocessed ring main unit connection part image to obtain a screw fastening key part image;
extracting key parts of the preprocessed ring main unit connection part image, wherein the key part extraction process comprises the following steps:
S41: calculating gradient directions of connected domain edge pixels in the preprocessed ring main unit continuous part images, wherein the connected domain edge pixels I' (I) * ,j * ) Gradient direction θ (i) * ,j * ) The calculation formula is as follows:
wherein:
g'(i * ,j * ) Representing connected domain edge pixels I' (I) * ,j * ) Is used for filtering the gray value after noise reduction;
s42: for connected domain edge pixels I' (I) * ,j * ) The coordinates of (i) are rotated to obtain rotated coordinates ((i) * )',(j * )')：
S43: affine transformation is performed on the rotated connected domain edge pixel coordinates, wherein the coordinates ((i) * )',(j * ) ') affine transformation formula:
wherein:
(i ', j') is the coordinate ((i) * )',(j * ) ') affine transformation results;
h is an affine transformation matrix;
s44: calculating Euclidean distance between each connected domain edge pixel coordinate and affine transformed coordinate, wherein the Euclidean distance between the D-th connected domain edge pixel coordinate and affine transformed coordinate is dis (D), the total number of the connected domain edge pixel coordinates is D, and the self-adaptive threshold is initialized to delta=0.1;
s45: marking the edge pixel coordinates of the connected domain with the Euclidean distance smaller than the self-adaptive threshold value as an inner point, otherwise marking as an outer point, and calculating the current evaluation value:
wherein:
n in representing the number of interior pointsOrder, n out Representing the number of outliers;
Ω in representing the set of interior points, Ω out Represents the outlier set, dis (d) in ) Representing the inner point d in Euclidean distance from affine transformed coordinates, dis (d out ) Representing the outer point d out Euclidean distance from affine transformed coordinates;
representing the average Euclidean distance between the edge pixel coordinates of the D connected domains and the affine transformed coordinates;
storing the current evaluation value and the interior point set;
s46: let δ=δ+0.3, return to step S45 until δ=3;
s47: traversing the currently stored evaluation value and the corresponding interior point set, and selecting the interior point set with the minimum evaluation value; regarding a plurality of connected domain images in the preprocessed ring main unit continuous part images, if the internal point proportion of the edge of the connected domain image reaches 30%, taking the connected domain image as a screw fastening key part image;
s5: and constructing a ring main unit fastening loosening identification model, wherein the constructed model takes an image of a key part of screw fastening as input and a screw loosening result as output, the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and if screw loosening exists, screw loosening early warning is carried out.
2. The method for rapid fault early warning of a multi-dimensional ring main unit according to claim 1, wherein the step S1 is to construct a ring main unit overheat index system, collect ring main unit overheat index data, and comprises:
Building a ring main unit overheat index system, wherein the ring main unit overheat index system is as follows:
(w 0 ,R,I,w)
wherein:
w 0 the environmental temperature index of the ring main unit is represented;
w represents an internal temperature index of the ring main unit;
r represents a contact resistance index of a bus connection point of the ring main unit;
i represents the index of the flowing current of the ring main unit;
the method comprises the steps of collecting ring main unit overheat index data based on a ring main unit overheat index system, wherein the collected ring main unit overheat index data are as follows:
{x(t)＝(x(t 0 ),x(t 1 ),...,x(t s ),...,x(t S ))|x∈{w 0 ,R,I,w}}
wherein:
x (t) represents time series data of the collected index x, t 0 Indicating the current time, t 1 ～t S The historical time is represented, and the time interval between adjacent times is Δt.
3. The method for fast warning of a failure of a multi-dimensional ring main unit according to claim 2, wherein the preprocessing of the ring main unit overheat index data in step S1 comprises:
preprocessing the ring main unit overheat index data, wherein the preprocessing flow is as follows:
s11: carrying out normalization processing on the acquired data of the index x at any moment, wherein the normalization processing formula is as follows:
wherein:
x(t s ) Representing t s Acquisition data of time index x, x' (t) s ) Represents x (t) s ) Is a normalization processing result of (a);
max x represents the maximum value, min, of the set index x x Representing the minimum value of the set index x;
S12: converting the normalized ring main unit overheat index data into a matrix form X:
wherein:
w 0 '(t S ),R'(t S ),I'(t S ),w'(t S ) Respectively t after normalization processing S Moment ring main unit overheat index w 0 Collecting data of R, I and w;
and taking X as a preprocessing result of the ring main unit overheat index data.
4. The method for rapid warning of multi-dimensional ring main unit faults as claimed in claim 3, wherein the step S2 of inputting the preprocessed ring main unit overheat index data into the ring main unit early warning model to obtain overheat risk early warning values comprises the following steps:
the method comprises the steps of constructing a ring main unit early warning model, wherein the ring main unit early warning model comprises an input layer, a characteristic calculation layer and a characteristic fusion layer, the input layer is used for receiving preprocessed ring main unit overheat index data, the characteristic calculation layer comprises three convolution layers and is used for carrying out convolution calculation on the preprocessed ring main unit overheat index data, the characteristic fusion layer is used for fusing calculation results of the three convolution layers to obtain ring main unit overheat index data at a prediction moment, and overheat risk early warning values are obtained based on ring main unit overheat index data at the prediction moment;
the preprocessed ring main unit overheat index data is input into a ring main unit early warning model to obtain overheat risk early warning values, and the overheat risk early warning value calculation flow based on the ring main unit early warning model is as follows:
The input layer of the ring main unit early warning model receives the preprocessed ring main unit overheat index data X;
setting the current iteration number as K, the initial value of K as 1, and the maximum iteration number as K, wherein the result obtained by the kth iteration is C k The result obtained by each iteration comprises the data of four ring main unit overheat indexes, namely ring main unit overheat index data at the predicted moment;
the iteration formula of the kth iteration is as follows:
C k ＝σ 2 ([X,C k-1 ]w 4 +b 4 )tanh(F 3,k-1 )
F 3,k-1 ＝σ 2 [(F 1,k-1 +F 2,k-1 )w 3 +b 3 ]
F 1,k-1 ＝σ 1 ([X,C k-1 ]w 1 +b 1 )
F 2,k-1 ＝σ 1 ([X,C k-1 ]w 2 +b 2 )
wherein:
w 1 ,w 2 ,w 3 weight parameters representing three convolutional layers in the feature computation layer, b 1 ,b 2 ,b 3 Representing bias parameters of three convolution layers in the feature calculation layer;
w 4 weight parameters representing feature fusion layer, b 4 Representing bias parameters of the feature fusion layer;
σ 1 (·),σ 2 (. Cndot.) represents an activation function; c (C) 0 The first column in X is the pre-processed ring main unit overheat index data at the current moment;
let k=k+1, repeat the iterative formula of the kth iteration until the iterative result of the kth iteration is obtained, namely the ring main unit overheat index data (w 0 * ,R * ,I * ,w * ) Wherein w is 0 * Ambient temperature index data representing the predicted time, R * Contact resistance index data representing ring main unit bus connection point at prediction time, I * Ring main unit flowing current index data w representing prediction time * The ring main unit internal temperature index data representing the prediction time;
ring main unit overheat index data (w) based on prediction time 0 * ,R * ,I * ,w * ) And calculating to obtain an overheat risk early warning value:
wherein:
and representing the overheat risk early warning value.
5. The method for rapid fault early warning of a multi-dimensional ring main unit according to claim 1, wherein if the early warning value of overheat risk exceeds a preset threshold value in step S3, collecting and preprocessing a continuous part image of the ring main unit, comprising:
if the overheat risk early warning value exceeds a preset threshold value, collecting and preprocessing a ring main unit connection part image, wherein the ring main unit connection part image is an image of a ring main unit screw connection position, and the preprocessing flow of the ring main unit connection part image is as follows:
s31: carrying out graying treatment on any pixel I (I, j) of the image of the ring main unit connection part, wherein the gray value of the pixel I (I, j) is as follows:
g(i,j)＝max{I R (i,j),I G (i,j),I B (i,j)}
wherein:
I R (i,j),I G (i,j),I B (I, j) are the color values of pixel I (I, j) in the RGB color channel, respectively;
g (I, j) represents the gray value of pixel I (I, j), and pixel I (I, j) represents the pixel of the ith row and the jth column in the image of the successive part of the ring main unit;
s32: filtering and noise reduction processing is carried out on the image of the ring main unit connection part obtained by the graying processing, and the filtering and noise reduction processing flow is as follows:
A gaussian kernel of 3×3 in size and 1 in standard deviation was set in the form of:
for any pixel in the image of the continuous part of the ring main unit after the graying treatment, multiplying a 3 multiplied by 3 pixel area taking the pixel as the center by a Gaussian kernel, and adding all numbers in the product result to obtain a filtering treatment result of the pixel, namely a gray value after the filtering treatment;
s33: for a 3×3 gray matrix a centered on an arbitrary pixel I (I, j), the Sobel operator S is used 1 And S is 2 Computing gradient matrix g for pixel I (I, j) 12 (i,j)：
Wherein:
* Representing a convolution process;
g 1 (I, j) represents a gradient matrix of the pixel I (I, j) in the horizontal direction;
g 2 (I, j) represents a gradient matrix of the pixel I (I, j) in the vertical direction;
s34: calculate g up (i, j) and g down (i, j) if g 12 The sum of the elements in (i, j) is greater than g up (i, j) and g down The sum of the elements of (I, j) then pixel I (I, j) is an edge pixel and the gray value of pixel I (I, j) is retained, otherwise it is set to 0, g up (i, j) and g down The calculation formula of (i, j) is:
wherein:
i represent L1 norm;
edge pixel detection is carried out on all pixels, so that an edge diagram of a connection part of the ring main unit is obtained;
s35: performing expansion corrosion operation on edge pixels in the edge map to obtain a plurality of connected domains, and calculating the area of each connected domain;
S36: deleting connected domain edge pixels with the connected domain area smaller than a preset area threshold, marking the connected domain edge pixel positions of the reserved connected domain in the ring main unit connected part image after filtering noise reduction processing, and taking the ring main unit connected part image after filtering noise reduction processing with the marked connected domain edge pixel positions as the ring main unit connected part image after preprocessing.
6. The method for rapid early warning of a multi-dimensional ring main unit fault according to claim 1, wherein the step S5 of obtaining the screw loosening result of the ring main unit by using the ring main unit tightening loosening recognition model comprises the following steps:
constructing a ring main unit fastening loosening identification model, wherein the constructed model takes a screw fastening key part image as input and a screw loosening result as output, and the ring main unit fastening loosening identification model is utilized to obtain the screw loosening result of the ring main unit, and the screw loosening result identification flow based on the ring main unit fastening loosening identification model is as follows:
s51: calculating to obtain the gradient direction of each pixel in the screw fastening key part image;
s52: and decomposing the image of the key part of screw fastening, wherein the decomposing formula is as follows:
wherein:
b represents a decomposition scale, C (B) represents a decomposition result under the scale B, wherein B E [1, B ], and B represents a maximum decomposition scale;
g '(h, y) represents a filtered and noise-reduced gray value of a pixel I' (h, y) in the screw tightening key part image;
θ (h, y) represents the gradient direction of the pixel I' (h, y);
s53: calculating to obtain maximum decomposition results C of B decomposition scales max And minimum decomposition result C min ；
S54: if it isIf the difference is smaller than the preset threshold value, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is not larger, the screw of the ring main unit is not loosened, otherwise, the difference of the decomposition result under the original scale 1 and the maximum decomposition result is larger, the screw of the ring main unit is loosened, and if the screw is loosened, screw loosening early warning is carried out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310777890.6A CN116822205B (en) | 2023-06-28 | 2023-06-28 | Rapid fault early warning method for multi-dimensional ring main unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310777890.6A CN116822205B (en) | 2023-06-28 | 2023-06-28 | Rapid fault early warning method for multi-dimensional ring main unit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116822205A CN116822205A (en) | 2023-09-29 |
| CN116822205B true CN116822205B (en) | 2023-12-29 |
Family
ID=88142609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310777890.6A Active CN116822205B (en) | 2023-06-28 | 2023-06-28 | Rapid fault early warning method for multi-dimensional ring main unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116822205B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104076242A (en) * | 2013-03-25 | 2014-10-01 | 北京电研华源电力技术有限公司 | Solid insulated ring network cabinet electrical joint detection method and device |
| CN113125996A (en) * | 2021-03-24 | 2021-07-16 | 河南工业职业技术学院 | Detection, regulation and control system of low-voltage electrical cabinet |
| CN113739923A (en) * | 2020-05-29 | 2021-12-03 | 比亚迪股份有限公司 | Power distribution cabinet fault monitoring method and system and vehicle |
-
2023
- 2023-06-28 CN CN202310777890.6A patent/CN116822205B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104076242A (en) * | 2013-03-25 | 2014-10-01 | 北京电研华源电力技术有限公司 | Solid insulated ring network cabinet electrical joint detection method and device |
| CN113739923A (en) * | 2020-05-29 | 2021-12-03 | 比亚迪股份有限公司 | Power distribution cabinet fault monitoring method and system and vehicle |
| CN113125996A (en) * | 2021-03-24 | 2021-07-16 | 河南工业职业技术学院 | Detection, regulation and control system of low-voltage electrical cabinet |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116822205A (en) | 2023-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113283446B (en) | Method and device for identifying object in image, electronic equipment and storage medium | |
| CN111652845A (en) | Abnormal cell automatic labeling method and device, electronic equipment and storage medium | |
| CN111695609A (en) | Target damage degree determination method, target damage degree determination device, electronic device, and storage medium | |
| CN116168351B (en) | Inspection method and device for power equipment | |
| CN112132216B (en) | Vehicle type recognition method and device, electronic equipment and storage medium | |
| CN113705462A (en) | Face recognition method and device, electronic equipment and computer readable storage medium | |
| CN111639704A (en) | Target identification method, device and computer readable storage medium | |
| US20200202514A1 (en) | Image analyzing method and electrical device | |
| CN116071520A (en) | Digital twin water affair simulation test method | |
| CN114494800B (en) | Predictive model training method and device, electronic equipment and storage medium | |
| CN114708461A (en) | Multi-modal learning model-based classification method, device, equipment and storage medium | |
| CN114049568A (en) | Object shape change detection method, device, equipment and medium based on image comparison | |
| CN113487621A (en) | Medical image grading method and device, electronic equipment and readable storage medium | |
| CN116822205B (en) | Rapid fault early warning method for multi-dimensional ring main unit | |
| CN115049836B (en) | Image segmentation method, device, equipment and storage medium | |
| CN115972198A (en) | Mechanical arm visual grabbing method and device under incomplete information condition | |
| CN113705686B (en) | Image classification method, device, electronic equipment and readable storage medium | |
| CN115147405A (en) | Rapid nondestructive testing method for new energy battery | |
| CN113963311A (en) | Safe production risk video monitoring method and system | |
| CN114387522A (en) | Intelligent early warning method, device, equipment and medium for working site | |
| CN112464946A (en) | Garbage image collection method and system based on big data | |
| CN117315368B (en) | Intelligent operation and maintenance inspection method for large-scale data center | |
| CN112580505A (en) | Method and device for identifying opening and closing states of network points, electronic equipment and storage medium | |
| CN114926753B (en) | Rapid target scene information extraction method under condition of massive images | |
| CN113160144B (en) | Target object detection method, target object detection device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |