Power equipment fault detection and diagnosis system

Abstract

The invention relates to the technical field of power equipment fault detection, in particular to a power equipment fault detection and diagnosis system, which comprises a preliminary detection and diagnosis module for collecting a panoramic image of power equipment in a region and completing preliminary detection and diagnosis of the power equipment according to an image data processing result; and the secondary detection and diagnosis module is used for carrying out secondary detection on the power equipment of which the fault cannot be judged through the temperature so as to finish accurate diagnosis of the fault. According to the method, heat source points and fault features are accurately identified by using gradient analysis of infrared image data and a support vector machine model; a histogram clipping and gray scale weighted fusion method is adopted, so that the influence of background noise on an image boundary is reduced, the contrast ratio between a heat source point and a background is enhanced, and the identification of a fault position is more accurate; on the basis of heat source point identification, multiple types of faults are classified, so that different types of faults such as short circuit, overload and grounding faults are efficiently and accurately classified.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment fault detection, and in particular to a power equipment fault detection and diagnosis system. Background Art

[0002] Power equipment fault detection and diagnosis is of vital importance to ensure the safe, stable and efficient operation of the power system. However, at present, the detection and diagnosis of power equipment faults faces the following challenges:

[0003] Traditional power equipment fault detection often relies on manual inspections, which have low inspection frequency and limited coverage. Inspection results are heavily dependent on the operator's experience and judgment, and there is a possibility of missed inspections or misjudgments. Due to the long inspection cycle, faults cannot be discovered in real time, which may cause minor faults to evolve into major accidents.

[0004] Traditional instruments focus more on the changes of single parameters such as current and voltage, and fail to fully capture the correlation of multi-dimensional data (such as temperature, vibration, power, etc.). For complex faults such as short circuit, grounding, and overload, traditional methods lack the ability to perform correlation analysis, resulting in inaccurate judgment of fault types.

[0005] Traditional methods are unable to collect and process large-scale equipment operation data in real time, making it difficult to promptly warn of potential failures. Inspection and measurement data are often simply recorded and fail to be deeply mined and analyzed, resulting in the inability to form effective historical trend forecasts. Summary of the invention

[0006] In view of the above-mentioned problems, the present invention is proposed.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: a power equipment fault detection and diagnosis system, comprising:

[0008] The preliminary detection and diagnosis module collects panoramic images of power equipment in the area and completes preliminary detection and diagnosis of power equipment based on the image data processing results.

[0009] The collected panoramic image data is divided into regional block grayscale levels, and the histogram is clipped according to the cumulative distribution function result of the regional block histogram. At the same time, the pixel values ​​between two regional blocks are weighted by using the adjacent regional block clipping results. According to the histogram clipping results and the weighted results, the preliminary detection and diagnosis of the power equipment is completed, which also includes:

[0010] According to the gradient of heat source image data, the contour of heat source points can be identified, the heat source area can be accurately determined, and the accuracy of preliminary detection and diagnosis of power equipment can be improved through the support vector machine model;

[0011] In addition, the secondary detection and diagnosis module performs secondary detection on power equipment that cannot be judged by temperature to complete accurate diagnosis of the fault, specifically:

[0012] Collect relevant working data of power equipment and construct multi-dimensional eigenvectors. Determine the covariance matrix based on eigenvalue calculations. At the same time, determine the correlation between eigenvalues ​​and fault types based on the covariance matrix to complete the determination of fault type classification. Based on the eigenvalues ​​corresponding to the fault types, accurate diagnosis of power equipment faults can be achieved.

[0013] As a preferred solution of the power equipment fault detection and diagnosis system of the present invention, the grayscale division of the collected panoramic image data into regional blocks is specifically as follows:

[0014] The image data I collected by the infrared sensor is converted into grayscale image data, and the grayscale image data is smoothed at the same time, then,

[0015] Select image data of size M×N, and divide the selected image data into area blocks of size k×k, and set the input panoramic image data pixel matrix I(x, y), where (x, y) represents the pixel value, and set the maximum pixel value I max And the minimum pixel value I min ;

[0016] For each area block in the image data, record the pixel gray value distribution histogram H in the area block i (g), where i represents the serial number of the region block and g represents the gray level.

[0017] As a preferred solution of the power equipment fault detection and diagnosis system described in the present invention, the histogram clipping according to the cumulative distribution function result of the regional block histogram is specifically as follows:

[0018] Set the maximum value of histogram pixel frequency to H max (g) = 0.01·k 2 , according to the set histogram pixel frequency, it is clipped, specifically:

[0019] If the pixel gray value distribution histogram corresponding to the region block satisfies the formula H i (g)>H max (g), the excess part is clipped and the clipped part is distributed to other gray levels;

[0020] Calculate the cumulative distribution function of each clipped histogram, then we have,

[0021]

[0022] Among them, j represents the serial number of the gray level, i represents the serial number of the region block, CDF represents the cumulative distribution function, CDF i (g) represents the cumulative distribution function result of the gray level of the i-th region block g, H i ′(g) represents the clipped histogram. If a clipped histogram is required, the value satisfies the formula H i ′(g)=H max (g) If the histogram does not need to be clipped, the value satisfies the formula H i ′(g)=H i (g);

[0023] The gray transformation of the pixels in the area block is performed according to the calculated cumulative function, specifically:

[0024]

[0025] Where I(x,y) represents the pixel matrix of panoramic image data, I max ,I min They represent the maximum pixel value and the minimum pixel value respectively, CDF represents the cumulative distribution function, i represents the serial number of the area block, and I′(x, y) represents the transformed pixel value.

[0026] As a preferred solution of the power equipment fault detection and diagnosis system of the present invention, the weighted processing of the pixel values ​​between two area blocks using the clipping results of adjacent area blocks is specifically as follows:

[0027] If the current pixel is located at the junction of two blocks, its grayscale value is weighted by the results of the four adjacent area blocks, specifically:

[0028] I fusion (x,y)=w1·I1′(x,y)+w2·I2′(x,y)+w3·I3′(x,y)+w4·I4′(x,y)

[0029] Among them, w1, w2, w3, and w4 represent the weights of the region block to the current pixel, which are determined by the distance between the center of the region block and the current pixel. I1′(x, y), I2′(x, y), I3′(x, y), and I4′(x, y) represent the pixel values ​​of the four region blocks after transformation. fusion (x, y) represents the pixel value after weighting;

[0030] Finally, all transformed pixel values ​​I′(x, y) and weighted pixel values ​​are combined to form the new heat source image data I out (x,y), which is used for subsequent preliminary fault detection of power equipment.

[0031] As a preferred solution of the power equipment fault detection and diagnosis system of the present invention, the preliminary detection and diagnosis of the power equipment is as follows:

[0032] The heat source image data I out T(x,y) in (x,y) is used as the temperature data at the pixel value (x,y) for heat source detection;

[0033] Set the upper limit of the power equipment fault judgment threshold T max And the standard operating temperature threshold T s By comparing the temperature data in the heat source image data with the power equipment fault judgment threshold, the preliminary detection and diagnosis of power equipment faults is completed.

[0034] If the comparison result satisfies the formula T(x,y)>T max , indicating that the temperature at the pixel value (x, y) exceeds the upper limit of the operating temperature of the power equipment, and the power equipment corresponding to the current area block fails;

[0035] If the comparison result satisfies the formula T(x,y)<T s , indicating that the temperature at the pixel value (x, y) is lower than the normal operating temperature of the power equipment, and the power equipment corresponding to the current area block has no abnormal fault;

[0036] If the comparison result satisfies the formula T s ≤T(x,y)≤T max , indicating that the temperature at the pixel value (x, y) is normal data, and the power equipment corresponding to the current area block is used as the detection data of the secondary detection and diagnosis module for secondary fault detection and diagnosis of the power equipment.

[0037] As a preferred solution of the power equipment fault detection and diagnosis system described in the present invention, the covariance matrix determined by eigenvalue calculation is specifically as follows:

[0038] According to the collected working status data of the power equipment, the real-time current and voltage deviation of the power equipment are monitored, then,

[0039] ΔI(t)=I(t)-I r

[0040] ΔV(t)=V(t)-V r

[0041] And construct a multidimensional feature vector F, and satisfy the formula,

[0042] F={I(t),V(t),ΔI(t),ΔV(t)}

[0043] Among them, I(t) represents real-time current data, V(t) represents real-time voltage data, I r Indicates the rated working current of the power equipment, V r Indicates the rated working voltage of the power equipment, ΔI(t) indicates the real-time current deviation, and ΔV(t) indicates the real-time voltage deviation;

[0044] Calculate the covariance matrix of the eigenvector F, then we have,

[0045]

[0046] Among them, ∑ represents the calculated covariance matrix, F i represents the i-th eigenvalue in the eigenvector, represents the mean of the eigenvector, T represents the matrix transpose, i represents the sequence number of the eigenvalue in the eigenvector, and n represents the upper limit of the eigenvalue types.

[0047] As a preferred solution of the power equipment fault detection and diagnosis system of the present invention, the fault type classification is determined as follows:

[0048]

[0049] Among them, W i represents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, Y represents the set of fault types, p(f i ,y) represents the joint distribution probability, which is F i =f i and the probability of Y=y occurring simultaneously, p(f i ), p(Y) represents the marginal probability distribution, respectively F i =f i and Y = the probability of y occurring alone;

[0050] According to the determined adaptive weights, the classification of power equipment fault types is completed, specifically:

[0051]

[0052] Among them, W i represents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, i represents the sequence number of the eigenvalue in the eigenvector, n represents the upper limit of the eigenvalue type, T ype Represents a set of power equipment fault types, including short circuit fault, overload fault, and ground fault.

[0053] As a preferred solution of the power equipment fault detection and diagnosis system of the present invention, the precise diagnosis of the power equipment fault is as follows:

[0054] Set the diagnostic thresholds corresponding to the fault types respectively, including the short-circuit current threshold I T,short , short circuit voltage threshold V T,short , overload time threshold T over And the ground resistance threshold R ground ;

[0055] At the same time, relevant data of power equipment during historical faults are introduced, including historical fault current data I his And historical fault voltage data V his ;

[0056] A multi-layer fault judgment mechanism is formulated to achieve accurate fault diagnosis of power equipment, specifically:

[0057] The diagnosis of short-circuit fault is based on the real-time current, real-time current, real-time current deviation and real-time voltage deviation in the characteristic vector, then,

[0058] When the real-time current instantaneous rise data and the real-time voltage instantaneous drop data satisfy the formula (I(t)>I T,short )∧(V(t)<V T,short ), the power equipment has the risk of short circuit failure, and the power equipment with this risk is subject to secondary inspection, specifically:

[0059] When the real-time current deviation data of the power equipment satisfies the formula ΔI>(I his -I r ), the real-time voltage data satisfies the formula ΔV<(V his -V r ), the fault diagnosis result of the current power equipment is a short circuit fault.

[0060] A computer device comprises a memory and a processor, wherein the memory stores a computer program and the processor implements the steps of a power equipment fault detection and diagnosis system when executing the computer program.

[0061] A computer-readable storage medium stores a computer program, which implements the steps of a power equipment fault detection and diagnosis system when executed by a processor.

[0062] Beneficial effects of the present invention:

[0063] The present invention uses the gradient analysis of infrared image data and the support vector machine model to accurately identify heat source points and fault characteristics;

[0064] The histogram clipping and grayscale weighted fusion methods are used to reduce the impact of background noise on the image boundary, enhance the contrast between the heat source point and the background, and make the identification of the fault location more accurate;

[0065] Based on the identification of hot source points, multi-class fault classification is performed, and the optimal classification hyperplane can be found in the high-dimensional space, so that different types of faults such as short circuit, overload and ground fault can be efficiently and accurately classified;

[0066] Through preliminary analysis of infrared image data, heat sources and changes in current and voltage can be identified to determine whether there are abnormal faults. The preliminary diagnosis module can quickly detect obvious faults in the preliminary screening. When the preliminary diagnosis cannot make a clear judgment, the system will enter the secondary diagnosis stage, further perform feature analysis based on current, voltage and other data, and combine machine learning algorithms such as support vector machines for classification to further refine the fault type. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0068] Figure 1 The present invention is a schematic diagram of the overall system step structure of a power equipment fault detection and diagnosis system. DETAILED DESCRIPTION

[0069] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0070] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0071] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0072] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0073] Meanwhile, in the description of the present invention, it should be noted that the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0074] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0075] Example 1

[0076] Reference Figure 1 , is an embodiment of the present invention, which provides a power equipment fault detection and diagnosis system, including a primary detection and diagnosis module, a secondary detection and diagnosis module and a feedback adjustment module,

[0077] Specifically, the preliminary detection and diagnosis module is used to perform preliminary detection and diagnosis on the power equipment in the area, and the secondary detection and diagnosis module is used to perform secondary detection and diagnosis on the power equipment in the area that has passed the preliminary detection, so as to ensure accurate diagnosis of power equipment faults.

[0078] Furthermore, the preliminary detection and diagnosis module collects panoramic images of power equipment in the area through infrared sensors, processes the collected panoramic image data of power equipment, and completes preliminary detection and diagnosis of power equipment faults according to the processing of image data. The specific implementation is as follows:

[0079] The infrared sensor is used to collect the panoramic image data I of the power equipment in the area, and a unique device number is set for the power equipment in the panoramic image data. Then, I={I1,I2,...,I N}, where I1 represents the first serial number of the power equipment in the panoramic image data, and N represents the total number of power equipment in the panoramic image data;

[0080] The collected panoramic image data is processed, specifically:

[0081] The image data I collected by the infrared sensor is converted into grayscale image data, and the grayscale image data is smoothed to improve the accuracy of preliminary fault detection and diagnosis of power equipment. Then,

[0082] Image data of size M×N is selected, and the selected image data is divided into k×k size regional blocks, the value of k is set by the implementer according to the actual application scenario, and in this embodiment, the value of k=8 is used for supplementary explanation, and the input panoramic image data pixel matrix I(x, y) is set, where (x, y) represents the pixel value, and the maximum pixel value I is set. max And the minimum pixel value I min ;

[0083] For each area block in the image data, record the pixel gray value distribution histogram H in the area block i (g), where i represents the serial number of the region block and g represents the gray level;

[0084] Clipping the histogram of each region block, we have,

[0085] Set the maximum value of histogram pixel frequency to H max (g) = 0.01·k 2 , according to the set histogram pixel frequency, it is clipped, specifically:

[0086] If the pixel gray value distribution histogram corresponding to the region block satisfies the formula H i (g)>H max (g), the excess part is clipped and the clipped part is distributed to other gray levels to ensure that all frequencies in the histogram do not exceed the maximum pixel frequency;

[0087] Calculate the cumulative distribution function of each clipped histogram, then we have,

[0088]

[0089] Among them, j represents the serial number of the gray level, i represents the serial number of the region block, CDF represents the cumulative distribution function, CDF i (g) represents the cumulative distribution function result of the gray level of the i-th region block g, H i ′(g) represents the clipped histogram. If a clipped histogram is required, the value satisfies the formula H i ′(g)=Hmax (g) If the histogram does not need to be clipped, the value satisfies the formula H i ′(g)=H i (g);

[0090] The gray transformation of the pixels in the area block is performed according to the calculated cumulative function, specifically:

[0091]

[0092] Where I(x,y) represents the pixel matrix of panoramic image data, I max ,I min They represent the maximum pixel value and the minimum pixel value respectively, CDF represents the cumulative distribution function, i represents the sequence number of the region block, and I′(x, y) represents the pixel value after transformation;

[0093] In order to eliminate the influence of the boundary area of ​​the regional block, the results between adjacent blocks are fused by setting the weight coefficient, then,

[0094] If the current pixel is located at the junction of two blocks, its grayscale value is weighted by the results of the four adjacent area blocks, specifically:

[0095] I fusion (x,y)=w1·I1′(x,y)+w2·I2′(x,y)+w3·I3′(x,y)+w4·I4′(x,y)

[0096] Among them, w1, w2, w3, and w4 represent the weights of the region block to the current pixel, which are determined by the distance between the center of the region block and the current pixel. I1′(x, y), I2′(x, y), I3′(x, y), and I4′(x, y) represent the pixel values ​​of the four region blocks after transformation. fusion (x, y) represents the pixel value after weighting;

[0097] Finally, all transformed pixel values ​​I′(x, y) and weighted pixel values ​​are combined to form the new heat source image data I out (x,y), which is used for subsequent preliminary fault detection of power equipment.

[0098] It should be noted that by setting the maximum value of the histogram pixel frequency, the histogram can suppress excessive enhancement of extreme pixel values ​​while locally enhancing it, trim the histogram high-frequency noise and fuse the blocks, thus avoiding the influence of boundary effects on the image.

[0099] Furthermore, the preliminary detection and diagnosis of the power equipment fault completed according to the processing of the image data is a new thermal source image data I out(x, y) performs heat source detection, and completes preliminary detection and diagnosis of power equipment faults based on the heat source detection results. The specific implementation is as follows:

[0100] The heat source image data I out T(x,y) in (x,y) is used as the temperature data at the pixel value (x,y) for heat source detection;

[0101] Set the upper limit of the power equipment fault judgment threshold T max And the standard operating temperature threshold T s By comparing the temperature data in the heat source image data with the power equipment fault judgment threshold, the preliminary detection and diagnosis of power equipment faults is completed.

[0102] If the comparison result satisfies the formula T(x,y)>T max , indicating that the temperature at the pixel value (x, y) exceeds the upper limit of the operating temperature of the power equipment, and the power equipment corresponding to the current area block fails;

[0103] If the comparison result satisfies the formula T(x,y)<T s , indicating that the temperature at the pixel value (x, y) is lower than the normal operating temperature of the power equipment, and the power equipment corresponding to the current area block has no abnormal fault;

[0104] If the comparison result satisfies the formula T s ≤T(x,y)≤T max , indicating that the temperature at the pixel value (x, y) is normal data, and the power equipment corresponding to the current area block is used as the detection data of the secondary detection and diagnosis module for secondary fault detection and diagnosis of the power equipment.

[0105] It should be noted that in order to identify the contour of the heat source point and prevent misjudgment caused by the blurred boundary of the high-temperature area, the gradient of the heat source image data is calculated, specifically:

[0106]

[0107] Among them, (x, y) represents the pixel value, I out (x,y) represents the heat source image data, The gradient of the calculated heat source image data is the temperature change at the pixel value (x, y), which is used to determine whether it is an edge area. Then,

[0108] Set the gradient judgment threshold T of the edge area R , and the initial temperature of the heat source point T0(x,y), and the detected heat source point (x0,y0) is used as the initial point to initialize the heat source point area R = {(x0,y0)};

[0109] Initially expand outward from the heat source point to complete the determination of the heat source area, then we have,

[0110] For the adjacent pixels (x′, y′) of the initial point (x0, y0) of the heat source, the temperature data T(x′, y′) of the adjacent pixels is collected, and the comparison of the gradient temperature data between the two satisfies the formula |T0(x, y)-T(x′, y′)|≤T R , then add the adjacent pixel point (x′, y′) to the heat source area R;

[0111] Repeat the heat source point expansion operation until all pixels that meet the conditions in the heat source area block have been added to the heat source area;

[0112] Extract the average temperature μ from the heat source region R R , maximum temperature T R,max And temperature variance data And construct the extracted data into a new feature vector

[0113] Finally, the constructed feature vector is used as the input data of the support vector machine model. According to the output results of the model, the heat source points are further classified. For the pixel points whose classification results are normal heat sources, the power equipment corresponding to the normal heat source pixel points is classified as normal working equipment. For the pixel points whose classification results are abnormal heat sources, the corresponding power equipment is classified as abnormal working equipment.

[0114] Furthermore, the secondary detection and diagnosis module is to perform secondary detection and diagnosis on the fault state of the power equipment that the preliminary detection and diagnosis module cannot judge according to the temperature, and complete the accurate diagnosis of the fault by constructing a feature vector, and setting an adaptive weight according to the correlation between the eigenvalue in the feature vector and the fault, and judging the fault type according to the adaptive weight, and completing the accurate fault diagnosis of the power equipment according to the eigenvalue corresponding to the fault type, which is specifically implemented as follows:

[0115] For power equipment whose fault status cannot be judged by temperature, collect the working status data of the power equipment, including real-time current data I(t) and real-time voltage data V(t), and record the rated working current I(t) of the power equipment. r And the rated working voltage V r ;

[0116] According to the collected working status data of the power equipment, the real-time current and voltage deviation of the power equipment are monitored, then,

[0117] ΔI(t)=I(t)-I r

[0118] ΔV(t)=V(t)-V r

[0119] And construct a multidimensional feature vector F, and satisfy the formula,

[0120] F={I(t),V(t),ΔI(t),ΔV(t)}

[0121] Among them, I(t) represents real-time current data, V(t) represents real-time voltage data, I r Indicates the rated working current of the power equipment, V r It represents the rated working voltage of the power equipment, ΔI(t) represents the real-time current deviation, and ΔV(t) represents the real-time voltage deviation.

[0122] Furthermore, the fault type adaptive weight is obtained by analyzing the correlation between the characteristic value and the fault, as follows:

[0123] Calculate the covariance matrix of the eigenvector F, then we have,

[0124]

[0125] Among them, ∑ represents the calculated covariance matrix, F i represents the i-th eigenvalue in the eigenvector, represents the mean of the eigenvector, T represents the matrix transpose, i represents the sequence number of the eigenvalue in the eigenvector, and n represents the upper limit of the eigenvalue types;

[0126] According to the covariance matrix, the correlation between the eigenvalue and the fault is calculated, and then the adaptive weight of the fault type is determined. Then,

[0127]

[0128] Among them, W i represents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, Y represents the set of fault types, p(f i ,y) represents the joint distribution probability, which is F i =f i and the probability of Y=y occurring simultaneously, p(f i ), p(Y) represents the marginal probability distribution, respectively F i =f i and Y = the probability of y occurring alone;

[0129] According to the determined adaptive weights, the classification of power equipment fault types is completed, specifically:

[0130]

[0131] Among them, W irepresents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, i represents the sequence number of the eigenvalue in the eigenvector, n represents the upper limit of the eigenvalue type, T ype Represents a set of power equipment fault types, including short circuit fault, overload fault and ground fault;

[0132] According to the characteristic value corresponding to the fault type, accurate fault diagnosis of power equipment is achieved, as follows:

[0133] Set the diagnostic thresholds corresponding to the fault types respectively, including the short-circuit current threshold I T,short , short circuit voltage threshold V T,short , overload time threshold T over And the ground resistance threshold R ground ;

[0134] At the same time, relevant data of power equipment during historical faults are introduced, including historical fault current data I his And historical fault voltage data V his ;

[0135] A multi-layer fault judgment mechanism is formulated to achieve accurate fault diagnosis of power equipment, specifically:

[0136] The diagnosis of short-circuit fault is based on the real-time current, real-time current, real-time current deviation and real-time voltage deviation in the characteristic vector, then,

[0137] When the real-time current instantaneous rise data and the real-time voltage instantaneous drop data satisfy the formula (I(t)>I T,short )∧(V(t)<V T,short ), the power equipment has the risk of short circuit failure, and the power equipment with this risk is subject to secondary inspection, specifically:

[0138] When the real-time current deviation data of the power equipment satisfies the formula ΔI>(I his -I r ), the real-time voltage data satisfies the formula ΔV<(V his -V r ), the fault diagnosis result of the current power equipment is a short circuit fault;

[0139] The diagnosis of overload fault is based on the real-time current, real-time current, real-time current deviation, real-time voltage deviation and duration in the characteristic vector, then,

[0140] When the real-time current data and real-time voltage data continue to rise within a unit time, and the duration t satisfies the formula t>T overWhen the power equipment is at risk of overload failure, a secondary inspection is performed on the power equipment with this failure risk, specifically:

[0141] When the real-time current deviation data of the power equipment satisfies the formula ΔI>(I his -I r ), the real-time voltage data satisfies the formula ΔV>(V his -V r ), the fault diagnosis result of the current power equipment is overload risk;

[0142] The diagnosis of ground fault is based on the real-time current data and ground resistance data in the characteristic vector, then,

[0143] When the zero-sequence current I0(t) in the real-time current data continues to rise within a unit time and is consistent with the grounding resistance current I ground (t) The comparison satisfies the formula I0(t)>I ground (t), the power equipment has the risk of ground fault, and the power equipment with this risk is subject to secondary inspection, specifically:

[0144] When the grounding resistance current of the power equipment satisfies the formula I0(t)>I ground (t), the ground impedance satisfies the formula R ground (t)<R ground When , the diagnosis result of the current power equipment is a ground fault.

[0145] It should be noted that for accurately diagnosed power equipment, the operating temperature data T m (t), and compare the collected working temperature data with the standard working temperature threshold T in the preliminary detection and diagnosis module s Compare, if the comparison result satisfies the formula T s (t)≥T s , it means that the standard operating temperature threshold data in the preliminary detection diagnosis module is low, and the setting of the standard operating temperature threshold is readjusted until the operating temperature data of all accurately diagnosed power equipment meets T s (t)<T s So far, it means that the set standard operating temperature threshold can ensure the accurate diagnosis of subsequent equipment on the basis of preliminary detection and diagnosis.

[0146] Furthermore, if the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the system described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0147] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0148] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0149] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0150] It will be appreciated that in the development of any actual implementation, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but will be a routine task of design, fabrication, and production for those of ordinary skill having the benefit of this disclosure without undue experimentation.

[0151] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A power equipment fault detection and diagnosis system, comprising: The preliminary detection and diagnosis module collects panoramic images of power equipment in the area and completes preliminary detection and diagnosis of power equipment based on the image data processing results. It is characterized by: The collected panoramic image data is divided into regional block grayscale levels, and the histogram is clipped according to the cumulative distribution function result of the regional block histogram. At the same time, the pixel values ​​between two regional blocks are weighted by using the adjacent regional block clipping results. According to the histogram clipping results and the weighted results, the preliminary detection and diagnosis of the power equipment is completed, which also includes: According to the gradient of heat source image data, the contour of heat source points can be identified, the heat source area can be accurately determined, and the accuracy of preliminary detection and diagnosis of power equipment can be improved through the support vector machine model; In addition, the secondary detection and diagnosis module performs secondary detection on power equipment that cannot be judged by temperature to complete accurate diagnosis of the fault, specifically: Collect relevant working data of power equipment and construct multi-dimensional eigenvectors. Determine the covariance matrix based on eigenvalue calculations. At the same time, determine the correlation between eigenvalues ​​and fault types based on the covariance matrix to complete the determination of fault type classification. Based on the eigenvalues ​​corresponding to the fault types, accurate diagnosis of power equipment faults can be achieved.

2. A power equipment fault detection and diagnosis system as claimed in claim 1, characterized in that: The grayscale division of the collected panoramic image data into regional blocks is specifically as follows: The image data I collected by the infrared sensor is converted into grayscale image data, and the grayscale image data is smoothed at the same time, then, Select image data of size M×N, and divide the selected image data into area blocks of size k×k, and set the input panoramic image data pixel matrix I(x, y), where (x, y) represents the pixel value, and set the maximum pixel value I max And the minimum pixel value I min ; For each area block in the image data, record the pixel gray value distribution histogram H in the area block i (g), where i represents the serial number of the region block and g represents the gray level.

3. A power equipment fault detection and diagnosis system as claimed in claim 2, characterized in that: The specific method of performing histogram clipping according to the cumulative distribution function result of the region block histogram is as follows: Set the maximum value of histogram pixel frequency to H max (g) = 0.01·k 2 , according to the set histogram pixel frequency, it is clipped, specifically: If the pixel gray value distribution histogram corresponding to the region block satisfies the formula H i (g)>H max (g), the excess part is clipped and the clipped part is distributed to other gray levels; Calculate the cumulative distribution function of each clipped histogram, then we have, Among them, j represents the serial number of the gray level, i represents the serial number of the region block, CDF represents the cumulative distribution function, CDF i (g) represents the cumulative distribution function result of the gray level of the i-th region block g, H i ′(g) represents the clipped histogram. If a clipped histogram is required, the value satisfies the formula H i ′(g)=H max (g) If the histogram does not need to be clipped, the value satisfies the formula H i ′(g)=H i (g); The gray transformation of the pixels in the area block is performed according to the calculated cumulative function, specifically: Where I(x,y) represents the pixel matrix of panoramic image data, I max ,I min They represent the maximum pixel value and the minimum pixel value respectively, CDF represents the cumulative distribution function, i represents the serial number of the area block, and I′(x, y) represents the transformed pixel value.

4. A power equipment fault detection and diagnosis system as claimed in claim 3, characterized in that: The weighted processing of the pixel values ​​between two area blocks using the clipping result of the adjacent area blocks is specifically as follows: If the current pixel is located at the junction of two blocks, its grayscale value is weighted by the results of the four adjacent area blocks, specifically: I fusion (x,y)=w1·I1′(x,y)+w2·I2′(x,y)+w3·I3′(x,y)+w4·I4′(x,y) Among them, w1, w2, w3, and w4 represent the weights of the region block to the current pixel, which are determined by the distance between the center of the region block and the current pixel. I1′(x, y), I2′(x, y), I3′(x, y), and I4′(x, y) represent the pixel values ​​of the four region blocks after transformation. fusion (x, y) represents the pixel value after weighting; Finally, all transformed pixel values ​​I′(x, y) and weighted pixel values ​​are combined to form the new heat source image data I out (x,y), which is used for subsequent preliminary fault detection of power equipment.

5. A power equipment fault detection and diagnosis system as claimed in claim 4, characterized in that: The preliminary detection and diagnosis of the power equipment are as follows: The heat source image data I out T(x,y) in (x,y) is used as the temperature data at the pixel value (x,y) for heat source detection; Set the upper limit of the power equipment fault judgment threshold T max And the standard operating temperature threshold T s By comparing the temperature data in the heat source image data with the power equipment fault judgment threshold, the preliminary detection and diagnosis of power equipment faults is completed. If the comparison result satisfies the formula T(x,y)>T max , indicating that the temperature at the pixel value (x, y) exceeds the upper limit of the operating temperature of the power equipment, and the power equipment corresponding to the current area block fails; If the comparison result satisfies the formula T(x,y)<T s , indicating that the temperature at the pixel value (x, y) is lower than the normal operating temperature of the power equipment, and the power equipment corresponding to the current area block has no abnormal fault; If the comparison result satisfies the formula T s ≤T(x,y)≤T max , indicating that the temperature at the pixel value (x, y) is normal data, and the power equipment corresponding to the current area block is used as the detection data of the secondary detection and diagnosis module for secondary fault detection and diagnosis of the power equipment.

6. A power equipment fault detection and diagnosis system as claimed in claim 5, characterized in that: The covariance matrix is ​​determined by calculating the eigenvalues ​​as follows: According to the collected working status data of the power equipment, the real-time current and voltage deviation of the power equipment are monitored, then, ΔI(t)=I(t)-I r ΔV(t)=V(t)-V r And construct a multidimensional feature vector F, and satisfy the formula, F={I(t),V(t),ΔI(t),ΔV(t)} Among them, I(t) represents real-time current data, V(t) represents real-time voltage data, I r Indicates the rated working current of the power equipment, V r Indicates the rated working voltage of the power equipment, ΔI(t) indicates the real-time current deviation, and ΔV(t) indicates the real-time voltage deviation; Calculate the covariance matrix of the eigenvector F, then we have, Among them, Σ represents the calculated covariance matrix, F i represents the i-th eigenvalue in the eigenvector, F represents the mean of the eigenvector, T represents the matrix transpose, i represents the sequence number of the eigenvalue in the eigenvector, and n represents the upper limit of the eigenvalue types.

7. A power equipment fault detection and diagnosis system as claimed in claim 6, characterized in that: The fault type classification is determined as follows: Among them, W i represents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, Y represents the set of fault types, p(f i ,y) represents the joint distribution probability, which is F i =f i and the probability of Y=y occurring simultaneously, p(f i ), p(Y) represents the marginal probability distribution, respectively F i =f i and Y = the probability of y occurring alone; According to the determined adaptive weights, the classification of power equipment fault types is completed, specifically: Among them, W i represents the adaptive weight corresponding to the i-th eigenvalue, F i represents the i-th eigenvalue in the eigenvector, i represents the sequence number of the eigenvalue in the eigenvector, n represents the upper limit of the eigenvalue type, T ype Represents a set of power equipment fault types, including short circuit fault, overload fault, and ground fault.

8. A power equipment fault detection and diagnosis system as claimed in claim 7, characterized in that: The precise diagnosis of the power equipment fault is as follows: Set the diagnostic thresholds corresponding to the fault types respectively, including the short-circuit current threshold I T,short , short circuit voltage threshold V T,short , overload time threshold T over And the ground resistance threshold R ground ; At the same time, relevant data of power equipment during historical faults are introduced, including historical fault current data I his And historical fault voltage data V his ; A multi-layer fault judgment mechanism is formulated to achieve accurate fault diagnosis of power equipment, specifically: The diagnosis of short-circuit fault is based on the real-time current, real-time current, real-time current deviation and real-time voltage deviation in the characteristic vector, then, When the real-time current instantaneous rise data and the real-time voltage instantaneous drop data satisfy the formula (I(t)>I T,short )∧(V(t)<V T,short ), the power equipment has the risk of short circuit failure, and the power equipment with this risk is subject to secondary inspection, specifically: When the real-time current deviation data of the power equipment satisfies the formula ΔI>(I his -I r ), the real-time voltage data satisfies the formula ΔV<(V his -V r ), the fault diagnosis result of the current power equipment is a short circuit fault.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the system according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the system according to any one of claims 1 to 8 are implemented.

Images