WO2020073201A1

WO2020073201A1 - Method and system for circuit breaker condition monitoring

Info

Publication number: WO2020073201A1
Application number: PCT/CN2018/109525
Authority: WO
Inventors: Niya CHEN; Jiayang RUAN; Rongrong Yu; Zhijian ZHUANG
Original assignee: Abb Schweiz Ag
Priority date: 2018-10-09
Filing date: 2018-10-09
Publication date: 2020-04-16
Also published as: EP3864424A1; EP3864424A4; CN111566493A; CN111566493B

Abstract

A method and system for circuit breaker condition monitoring are disclosed. The method comprises: obtaining an image of a circuit breaker(202); extracting from the image one or more features related to a state of the circuit breaker(204); comparing the extracted one or more features with benchmark data characterizing a predetermined state of the circuit breaker(206); determining a health condition of the circuit breaker based on the comparison(208).

Description

METHOD AND SYSTEM FOR CIRCUIT BREAKER CONDITION MONITORING

TECHNICAL FIELD

Example embodiments of the present disclosure generally relate to circuit breaker monitoring and more particularly, to a method and system for circuit breaker condition monitoring.

BACKGROUND

Circuit breakers are widely used in an electrical grid. Circuit breakers are designed to protect an electrical circuit or electrical devices from damage caused by excess current from an overload or short circuit. When circuit breakers fail to operate during such an adverse electrical condition, catastrophic results may arise. However, the circuit breakers may be subject to various failures over time, which will threaten security of the electrical circuit. It is desirable to carry out condition monitoring of the circuit breakers so as to track the operation conditions of the circuit breakers and to enable the indication of potential failure occurrences and preventive maintenance.

Circuit breakers are generally enclosed in a casing and their conditions cannot be easily monitored. Conventional circuit breaker monitoring systems typically comprises a measuring device that measures parameters associated with the circuit breaker. However, such a system can only monitor limited states of circuit breaker and cannot provide a comprehensive condition monitoring and diagnosis of the circuit breakers.

In some circuit breakers, a camera is provided to take pictures of the circuit breaker. However, the pictures taken are typically transmitted to a remote control center where technical engineers check and diagnose the pictures taken one by one to carry out condition monitoring. This is time-consuming and inefficient.

Additionally, in some circuit breakers, a machine vision method is used, but it is irrelative to condition monitoring and diagnosis of the circuit breakers. For example, CN106526467A discloses a high voltage circuit breaker which can measure a closing and opening speeds of the movable contact based on machine vision. A reference point is predetermined, and positions of an object in a series of images are identified and the closing and opening speeds of the movable contact then can be calculated.

SUMMARY

Example embodiments of the present disclosure propose a solution for circuit breaker condition monitoring.

In a first aspect, example embodiments of the present disclosure provide a method for circuit breaker condition monitoring. The method comprises: obtaining an image of a circuit breaker; extracting from the image one or more features related to a state of the circuit breaker; comparing the extracted one or more features with benchmark data characterizing a predetermined state of the circuit breaker; determining a health condition of the circuit breaker based on the comparison.

With the method for circuit breaker condition monitoring, the health condition of the circuit breaker can be reliably and accurately determined in a simply way. Moreover, since the condition monitoring of the circuit breakers is carried out by machine vision method, any number of predetermined states of the circuit breaker can be monitored without adding substantial burden. Accordingly, some unhealthy conditions such as conductor corrosion, dust pollution, which are difficult to be monitored using conventional methods, can be monitored in a simple way.

In some embodiments, extracting the one or more features comprises at least one of the following: obtaining a binary image or a grayscale histogram of the image; extracting from the image key feature points describing features of components of the circuit breaker; and segmenting the image to identify positions and grayscale data of components of the circuit breaker. Accordingly, computation complexity is simplified.

In some embodiments, the components of the circuit breaker comprise at least one of the following: a stationary contact, a movable contact, and a grounding contact.

In some embodiments, the comparing comprises: determining a metric including at least one of the following: a distance between the one or more features and the benchmark data, and a correlation coefficient between one or more features and the benchmark data; and determining a dissimilarity between the one or more features and the benchmark data based on the metric. Accordingly, the health condition of the circuit breaker can be determined by image processing methods.

In some embodiments, the method further comprises: obtaining groups of images, each group including a plurality of images for one of predetermined states of the circuit breaker; for each group of images, extracting features characterizing the respective predetermined state from the plurality of images; and training a classifier based on the extracted features. When a group of images are provided for one of predetermined states of the circuit breaker, the classifier can be well trained. The reliability and accuracy of determination can be improved.

In some embodiments, the one or more extracted features are classified by the classifier. When the classifier is used, the computation is reduced and the health condition of the circuit breaker can be determined in a convenient and efficient manner.

In some embodiments, the predetermined states comprise at least one of the following: normal closed, defective closed, normal opening, defective opening, normal grounding, defective grounding, conductor corrosion, and dust pollution.

In a second aspect, example embodiments of the present disclosure a system for circuit breaker condition monitoring. The system comprises: a camera configured to take a picture of the circuit breaker; and at least one processor communicatively coupled to the camera and configured to perform the method of the first aspect. In some embodiments, the at least one processor may be local. In some embodiments, the at least one processor may be remote. With the system for circuit breaker condition monitoring, the health condition of the circuit breaker can be reliably and accurately determined in a simply way. All advantages with regard to the method can be analogously achieved.

In a third aspect, example embodiments of the present disclosure provide a computer readable medium having instructions stored thereon, the instructions, when executed on at least one processor, cause the at least one processor to perform the method according to any of the first aspect.

In a fourth aspect, example embodiments of the present disclosure provide a computer program product being tangibly stored on a computer readable storage medium and comprising instructions which, when executed on at least one processor, cause the at least one processor to perform the method according to any of the first aspect.

In a fifth aspect, example embodiments of the present disclosure provide an Internet of Things (IoT) system. The system comprise: a circuit breaker; and a system for circuit breaker condition monitoring according the second aspect.

BRIEF DESCRIPTION OF THE DRAWINGS

Through the following detailed descriptions with reference to the accompanying drawings, the above and other objectives, features and advantages of the example embodiments disclosed herein will become more comprehensible. In the drawings, several example embodiments disclosed herein will be illustrated in an example and in a non-limiting manner, wherein:

Figs. 1 illustrate a schematic view of a circuit breaker in accordance with some example embodiments of the present disclosure;

Fig. 2 illustrates a flowchart of a method for circuit breaker condition monitoring in accordance with some example embodiments of the present disclosure;

Fig. 3 illustrates a schematic diagram of feature extraction processes in accordance with some example embodiments of the present disclosure;

Fig. 4 illustrates images showing various predetermined states of the circuit breaker in accordance with some example embodiments of the present disclosure;

Fig. 5 illustrates grayscale histograms corresponding to the images in Fig. 4 in accordance with some example embodiments of the present disclosure;

Fig. 6 illustrates two dimensional feature vectors extracted from images of the circuit breaker in accordance with some example embodiments of the present disclosure; and

Fig. 7 illustrates a test image and its grayscale histogram in accordance with some example embodiments of the present disclosure.

Fig. 8 illustrates a block diagram of a system 800 for circuit breaker condition monitoring in accordance with embodiments of the present disclosure.

Throughout the drawings, the same or corresponding reference symbols refer to the same or corresponding parts.

DETAILED DESCRIPTION

The subject matter described herein will now be discussed with reference to several example embodiments. These embodiments are discussed only for the purpose of enabling those skilled persons in the art to better understand and thus implement the subject matter described herein, rather than suggesting any limitations on the scope of the subject matter.

The term “comprises” or “includes” and its variants are to be read as open terms that mean “includes, but is not limited to. ” The term “or” is to be read as “and/or” unless the context clearly indicates otherwise. The term “based on” is to be read as “based at least in part on. ” The term “being operable to” is to mean a function, an action, a motion or a state can be achieved by an operation induced by a user or an external mechanism. The term “one embodiment” and “an embodiment” are to be read as “at least one embodiment. ” The term “another embodiment” is to be read as “at least one other embodiment. ”

Unless specified or limited otherwise, the terms “mounted, ” “connected, ” “supported, ” and “coupled” and variations thereof are used broadly and encompass direct and indirect mountings, connections, supports, and couplings. Furthermore, “connected” and “coupled” are not restricted to physical or mechanical connections or couplings. In the description below, like reference numerals and labels are used to describe the same, similar or corresponding parts in the Figures. Other definitions, explicit and implicit, may be included below.

Figs. 1 illustrate a schematic view of a circuit breaker in accordance with some example embodiments of the present disclosure. The view is merely shows its main components, and other components are omitted.

As shown, the circuit breaker 10 comprises a fixed contact 14, a movable contact 12, and a ground contact 16. The movable contact 12 may include many operation states. For example, the movable contract 12 may move to a first position where it contacts the fixed contact 14. In this case, the circuit including the circuit breaker 10 is closed and current can flow thorough the circuit breaker 10. The movable contract 12 may move to a second position where it separates from the fixed contact 14 and the ground contact 16. In this case, the circuit including the circuit breaker 10 is open and current cannot flow thorough the circuit breaker 10. The movable contract 12 may also move to a third position where it contacts the ground contact 16, so that the circuit breaker 10 is grounded so as to protect the circuit breaker 10 from damage.

The circuit breaker 10 is important for the circuit protection and thus its operation states should be well monitored. As mentioned above, some conventional condition monitoring methods can only detect a limited number of states of the circuit breaker, and some conventional condition monitoring methods deeply need human participation and cannot be automatically carried out, which is inefficient and inconvenient. The present disclosure proposes a novel solution for circuit breaker condition monitoring and diagnosis based on image processing or analyzing techniques.

Fig. 2 illustrates a flowchart of a method for circuit breaker condition monitoring 200 in accordance with some example embodiments of the present disclosure. The method 200 can be implemented to carry out condition monitoring of the circuit breaker efficiently in a simply way and without any influence on the operations of the circuit breaker.

At block 202, one or more images of a circuit breaker are taken. At least one camera or the like can be fixedly amounted at proper positions within the circuit breaker. It is preferred that the position of the camera can ensure that the images taken by the camera include the key features for condition monitoring or diagnosis of the circuit breaker. In some embodiments, only one camera is provided which can meet the requirements of condition monitoring. In some embodiments, two or more cameras are provided so as to more data for condition monitoring.

At block 204, one or more features related to a state of the circuit breaker are extracted from the image. Generally, the pictures that are taken by the camera are colored. In some embodiments, these pictures are recorded as 3 matrices in computer. The data are generally very big and is hard to be directly used. The pictures have to be compressed or processed to extract key or essential features that are associated with condition monitoring or diagnosis of the circuit breaker. There are many methods for extracting features from the images, which will be described below, for example with reference to Figs. 3 and 4.

At block 206, the extracted one or more features are compared with benchmark data characterizing a predetermined state of the circuit breaker. In the method 200 in accordance with some example embodiments of the present disclosure, benchmark data are created for characterizing predetermined states of the circuit breaker. In some embodiments, these benchmark data are stored in a database accessible to a processor. Before benchmark data are created, data related to predetermined states of the circuit breaker have to be collected. For example, the images which record the predetermined states of the circuit breaker are collected, and these collected images are processed to extract key features or essential features. These extracted features are used as benchmark data for characterizing a predetermined state of the circuit breaker. In some embodiments, the method can determine at least one predetermined state of the circuit breaker When a plurality of predetermined state of the circuit breaker are to be determined, for each predetermined state, at least one image which records or describes the predetermined states of the circuit breaker is collected to form the benchmark data.

At block 208, a health condition of the circuit breaker can be determined based on the comparison. When the health condition of the circuit breaker is determined, the health condition, for example, is displayed to the user via a display. In some embodiments, when the system comprises a number of circuit breakers, the states of the circuit breakers can be reviewed at the same time. In some embodiments, when the condition of the circuit breaker is unhealthy, an alarmed is sent to the user, and proper actions, for example, replacement or maintenance of the circuit breaker, can be taken.

According to the method for circuit breaker condition monitoring, the health condition of the circuit breaker can be reliably and accurately determined in a simply way. Moreover, since the condition monitoring of the circuit breakers is carried out by machine vision method, any number of predetermined states of the circuit breaker can be monitored without adding substantial burden. Accordingly, some unhealthy conditions such as conductor corrosion, dust pollution, which are difficult to be monitored using conventional methods, can be monitored in a simple way.

In some embodiments, the pictures taken by the camera typically can be compressed or processed to extract key or essential features. There are many methods for extracting features from the images. When the pictures are further processed, computation complexity is simplified and the burden on the hardware can be reduced.

Fig. 3 illustrates schematic diagrams of feature extraction in accordance with some example embodiments of the present disclosure. As shown in Fig. 3, the original colored image 301 taken by the camera is transferred to a grayscale image 302. Since a grayscale image is typically recorded by 1 matrix, the original colored image 301 can be compressed.

In the embodiments shown, the greyscale image can be further compressed to reduce computation complexity of image processing.

In one example embodiment, the grayscale image 302 can be transferred to a binary image 303 using various binaryzation methods. As an example, an Otsu method is used for binaryzation. In Otsu′s method the threshold that minimizes an intra-class variance (i.e., variance within the class) , is exhaustively searched for. The intra-class variance is defined as a weighted sum of variances of the two classes:

where weights ω ₀ and ω ₁ represent the probabilities of the two classes separated by a threshold t, respectively, and

and

represent variances of these two classes, respectively.

In another example embodiment, the grayscale image 302 can be transferred to a grayscale histogram 304. A grayscale histogram is a type of histogram that acts as a graphical representation of the grayscale distribution in a digital image. It plots the number of pixels for each grayscale value. The grayscale histogram for a specific image records the grayscale distribution of the image. The grayscale histogram vector can then be used as vector features for calculation.

In some example embodiments, feature points in image are extracted using image processing algorisms. As an example, interest of points on the objects in image can be extracted. These features may characterize the essential features of the circuit breaker. The characterized features can be used to identify the key objects in different images can be used to compare with the benchmark data. As noted, to perform reliable recognition, the features extracted from the images be detectable even under changes in noise and illumination. For example, such points typically lie on high-contrast regions of the image, such as object edges. In some example embodiments, these features include edges, corners, or the like of the components in the image. The image processing algorisms may include SIFT (Scale-invariant feature transform) , corner detection, etc.

In some example embodiments, image segmentation method is used to locate key components. A location and surface state of key components can reflect the state of the circuit breaker. When the location and the surface state of key components in the image are identified, this information can be used to determine the states of the circuit breaker. For example, the location of a movable contact can be used to determine the open and closed states of the circuit breaker, the erosion feature in copper contacts can represent defective characteristics. Typically, domain-knowledge database can be created in advance. Image segmentation methods can be used in combination with domain-knowledge to locate the key components and abstract the corresponding part in figure as input for diagnosis model. As an example, a movable contact can be localized by image segmentation in combination with domain knowledge as a circle inside a rectangle. The location and surface state of a movable contact can then be used for the circuit breaker state determination. Image segmentation method can be carried out in various algorithms, including but not limited to K-means, Watershed, GraphCut, etc.

With reference to Figs. 4-6, a method for creating the benchmark data is described. For example, images related to predetermined states of the circuit breaker are collected. Benchmark data characterizing the key features associated the predetermined states of the circuit breaker are extracted from these images using the extracting methods mentioned above. These benchmark data are then stored in the database. The predetermined states comprise but not limited to normal closed, defective closed, normal opening, defective opening, normal grounding, defective grounding, conductor corrosion, and dust pollution.

In one example, Fig. 4 merely illustrates four predetermined states of the circuit breaker in accordance with some example embodiments of the present disclosure. As shown at 401 in Fig. 4, the circuit breaker is closed, and the movable contact contacts the fixed contract. At 402 in Fig. 4, the circuit breaker is grounded, and the movable contact contacts the grounding contract. At 403 in Fig. 4, the circuit breaker is open, and the movable contact is located between the fixed contact and the grounding contract and does not contact any of them. At 404 in Fig. 4, the circuit breaker is defectively closed, and the movable contact is close to or partially contacts the fixed contract.

The collected images that represent the predetermined states of the circuit breaker can be compressed to extract key features for characterizing the predetermined states.

Fig. 5 illustrates grayscale histograms corresponding to the images in Fig. 4 in accordance with some example embodiments of the present disclosure. The grayscale histograms associated with the predetermined states of the circuit breaker can be stored, as benchmark data, in the database. As shown in Fig. 5, the grayscale histogram 501 is corresponding to the closing state of the circuit breaker in Fig. 4; the grayscale histogram 502 is corresponding to the grounding state of the circuit breaker in Fig. 4; the grayscale histogram 503 is corresponding to the opening state of the circuit breaker in Fig. 4; and the grayscale histogram 504 is corresponding to the defective closing state of the circuit breaker in Fig. 4.

Fig. 6 illustrates two dimensional feature vectors extracted from images of the circuit breaker in accordance with some example embodiments of the present disclosure. Since the vector is two dimensional, it can be represented as a dot in two-dimensional coordinate frame. As shown, each dot in the figure represents feature vector extracted from one figure of the circuit breaker. Different types of benchmark data are collected as reference data. In the shown example, three types of known normal circuit breaker state and a possible defective closing type are shown.

As shown, the reference signs 601 ( “×” ) represents that the circuit breaker can be normally opened. The reference signs 602 ( “△” ) represents that the circuit breaker can be normally closed. The reference signs 603 “☆” represents that the circuit breaker can be normally grounded. The reference signs 604 “○” represents one type of defectiveness. For example, the circuit breaker cannot be normally closed or opened, i.e., defective opening or defective closing. The reference signs 605

represents another type of defectiveness. For example, the circuit breaker cannot be normally grounded, i.e, defective grounding. The reference signs 606 “*” represents the extracted test feature vector. As shown, as for each signs of benchmark data, a plurality of points is shown. That means, there is a plurality of images which are collected to indicate each predetermined state. In some embodiments, when only one image for each predetermined state is collected, only one dot is shown for each predetermined state. Test feature vector can be classified based on benchmark data using multiple methods, e.g. K-nearest neighbor, support vector machine, classical image processing methods such as similarity calculation, and the like.

In order to determine the state of a target circuit breaker, a test image or picture is taken by a camera and features for characterizing the state of the target circuit breaker are extracted. The extracted test features or vector can be classified based on benchmark data using multiple methods. The methods include but are not limited to classical image processing methods and machine learning algorisms.

Now embodiments of the method for determining the state of a target circuit breaker will be described. The main principle of these methods is to determine to which type of benchmark the test data is the most similar.

In some example embodiments, classical image processing methods are used. For example, a metric, such as a distance between the one or more features and the benchmark data, and a correlation coefficient between one or more features and the benchmark data is determined, and dissimilarity between the one or more features and the benchmark data based on the metric is determined. Then, the state of a target circuit breaker can be determined.

Fig. 7 illustrates a test image 701 and its grayscale histogram 702 in accordance with some example embodiments of the present disclosure. The original colored image 701 taken by the camera is transferred to a grayscale histogram 702 using the above mentioned method. In some example embodiments, the grayscale histograms shown in Fig. 5 are stored as benchmark data.

Grayscale histogram is recorded as one feature vector, for example, including 256 elements, corresponding to the number of pixels in the image at each grayscale value (grayscale = 1～256) . It is to be understood that length of grayscale histogram vector does not have to be 256, the length can be determined as other values. Accordingly, the benchmark data shown in Fig. 5 represent four classes i.e., predetermined states of the circuit breaker, each denoted as benchmark feature vector expressed by x _j (j = 1, 2, 3, 4) , and the extracted feature vector of the test image is denoted as x _t.

In some embodiments, dissimilarity or/and distance may be calculated so as to classify the test image as one class in the benchmark data. The benchmark feature vectors to which the feature vector of the test image is most similar or the nearest can indicate the state of the circuit breaker. As an example, the Euclidean distance d _jbetween feature vector of test image and each benchmark feature vector can be calculated using the following equation.

The results Euclidean distance from the feature vector of test image to each benchmark feature vector are in table 1 as below:

TABLE 1

It can be seen from above that d3 is the minimum and therefore the test image can be classified or determined as “OPEN” .

There is a plurality of ways of calculating dissimilarity or distance. For example, in some example embodiments, the correlation coefficient between feature vector of test image and each benchmark feature vector can be calculated.

In some example embodiments, machine learning algorisms are used. For example, in one embodiment, groups of images, each group including a plurality of images for one of predetermined states of the circuit breaker, are obtained. For each group of images, features characterizing the respective predetermined state from the plurality of images are extracted. A classifier based on the extracted features is trained. Then, the state of a target circuit breaker can be determined by the classifier. When a group of images are provided for one of predetermined states of the circuit breaker, the classifier can be well trained. This can improve the reliability and accuracy of determination. There is a plurality of machine learning methods can be used to train a classifier based on benchmark data, e.g. K Nearest Neighbor (KNN) , Support Vector Machine (SVM) , neural network, logistic regression, etc. As an example, the principle of classification using KNN and SVM methods are given here as below. It is to be understood the following examples are merely illustrative.

In a KNN method, the input is the extracted test feature, and the output is a class membership. The test image taken by the camera is classified by a majority vote of its neighbors in the benchmark database and the test data is assigned to the class most common among its k nearest neighbors (k is a positive integer, typically small) .

It is assumed that the benchmark data is denoted as (X ₁, y ₁) , (X ₂, y ₂) , ..., (X _n, y _n) , where the state feature is denoted as X _i, and y _i is the state class of the state feature X _i. Given distance definition (e.g. Euclidean distance) , the nearest k samples in benchmark dataset can be determined. The majority vote of these k samples indicates the state or the class of test image.

In some embodiments, a threshold can be set for classification. That is, if the distance between the test feature vector and the nearest benchmark neighbor is larger than the threshold, the state of test image can be determined as “other” , meaning that no known benchmark is similar to test image. This may be a new defective type. In some embodiments, in this event, an alert can be sent to the user.

In some example embodiments, a SVM method may be used to train the classifier model can be trained based on, creating a representation of the benchmark points that the separate categories are divided by a clear gap that is as wide as possible.

For example, given benchmark data (X ₁, y ₁) , (X ₂, y ₂) ..., (X _n, y _n) , where the state feature is denoted as X _i, and y _i is the state class of the state feature X _i. The target of training using SVM is to obtain a classifier or a model.

that can minimize loss function R [f]

subject to

where

represents a mapping function from a lower dimensional space to a higher dimensional space, w represents weight of the mapping function, and b represents bias. By the loss function R [f] in equation (4) and constraints in equation (5) , the value of w in equation (3) can be determined. Thus, the classifier or a model can be obtained in equation (3) .

Accordingly, for any test data x _*, the state of the circuit breaker is the calculated output y _*by trained classifier.

Fig. 8 shows a block diagram of a system 800 for circuit breaker condition monitoring in accordance with embodiments of the present disclosure. The system 800 comprises a camera 805 and at least one processor 810. The camera 805 is configured to take a picture of the circuit breaker. The at least one processor 810 is communicatively coupled to the camera 805 and configured to perform the method 200 as described above. With the system for circuit breaker condition monitoring, the health condition of the circuit breaker can be reliably and accurately determined in a simply way. All advantages with regard to the method 200 can be analogously achieved, which will not be repeatedly described herein.

Generally, various embodiments of the present disclosure may be implemented in hardware or special purpose circuits, software, logic or any combination thereof. Some aspects may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device. While various aspects of embodiments of the present disclosure are illustrated and described as block diagrams, flowcharts, or using some other pictorial representation, it will be appreciated that the blocks, apparatus, systems, techniques or methods described herein may be implemented in, as non-limiting examples, hardware, software, firmware, special purpose circuits or logic, general purpose hardware or controller or other computing devices, or some combination thereof.

The present disclosure also provides at least one computer program product tangibly stored on a non-transitory computer readable storage medium. The computer program product includes computer-executable instructions, such as those included in program modules, being executed in a device on a target real or virtual processor, to carry out the process or method as described above with reference to Fig. 2. Generally, program modules include routines, programs, libraries, objects, classes, components, data structures, or the like that perform particular tasks or implement particular abstract data types. The functionality of the program modules may be combined or split between program modules as desired in various embodiments. Machine-executable instructions for program modules may be executed within a local or distributed device. In a distributed device, program modules may be located in both local and remote storage media.

Program code for carrying out methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowcharts and/or block diagrams to be implemented. The program code may execute entirely on a machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

The above program code may be embodied on a machine readable medium, which may be any tangible medium that may contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine readable medium may be a machine readable signal medium or a machine readable storage medium. A machine readable medium may include but not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine readable storage medium would include an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM) , a read-only memory (ROM) , an erasable programmable read-only memory (EPROM or Flash memory) , an optical fiber, a portable compact disc read-only memory (CD-ROM) , an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

Further, while operations are depicted in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Likewise, while several specific implementation details are contained in the above discussions, these should not be construed as limitations on the scope of the present disclosure, but rather as descriptions of features that may be specific to particular embodiments. Certain features that are described in the context of separate embodiments may also be implemented in combination in a single embodiment. On the other hand, various features that are described in the context of a single embodiment may also be implemented in multiple embodiments separately or in any suitable sub-combination.

Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are disclosed as example forms of implementing the claims.

Claims

A method for circuit breaker condition monitoring, comprising:

obtaining an image of a circuit breaker;

extracting from the image one or more features related to a state of the circuit breaker;

comparing the extracted one or more features with benchmark data characterizing a predetermined state of the circuit breaker;

determining a health condition of the circuit breaker based on the comparison.
The method according to claim 1, wherein extracting the one or more features comprises at least one of the following:

obtaining a binary image or a grayscale histogram of the image;

extracting from the image key feature points describing features of components of the circuit breaker; and

segmenting the image to identify positions and grayscale data of components of the circuit breaker.
The method according to any one of preceding claims, wherein the components of the circuit breaker comprise at least one of the following: a stationary contact, a movable contact, and a grounding contact.
The method according to any one of preceding claims, wherein the comparing comprises:

determining a metric including at least one of the following: a distance between the one or more features and the benchmark data, and a correlation coefficient between one or more features and the benchmark data; and

determining a dissimilarity between the one or more features and the benchmark data based on the metric.
The method according to any one of preceding claims, further comprising:

obtaining groups of images, each group including a plurality of images for one of predetermined states of the circuit breaker; and

for each group of images,

extracting features characterizing the respective predetermined state from the plurality of images; and

training a classifier based on the extracted features.
The method according to claim 5, wherein the one or more extracted features are classified by the classifier.
The method according to any of the preceding claims, wherein the predetermined states comprise at least one of the following: normal closed, defective closed, normal opening, defective opening, normal grounding, defective grounding, conductor corrosion, and dust pollution.
A system for circuit breaker condition monitoring comprising:

a camera configured to take a picture of the circuit breaker; and

at least one processor communicatively coupled to the camera and configured to perform the method of any of claims 1 to 7.
A computer readable medium having instructions stored thereon, the instructions, when executed on at least one processor, cause the at least one processor to perform the method according to any of claims 1 to 7.
A computer program product being tangibly stored on a computer readable storage medium and comprising instructions which, when executed on at least one processor, cause the at least one processor to perform the method according to any of claims 1 to 7.
An Intemet of Things (IoT) system comprising:

a circuit breaker; and

a system for circuit breaker condition monitoring according to claim 8.