Method and device for identifying position of circuit breaker trolley in medium-voltage switch cabinet

Abstract

The invention relates to a method and device for identifying the position of a circuit breaker trolley in a medium-voltage switch cabinet, and the method comprises the steps: obtaining a plurality of continuous images of the circuit breaker trolley, calculating the distance between each continuous image and a reference image, analyzing the trend of the calculated distance, and determining that the circuit breaker trolley is not at a working position when the trend indicates that the circuit breaker trolley is moving. When the trend indicates that the circuit breaker trolley is stable, a first convolutional neural network model is used for detecting the existence of a plum blossom contact in the latest image, when the plum blossom contact is detected, it is determined that the circuit breaker trolley is not in the working position, and when the plum blossom contact is not detected, the distance between the latest image and the reference image is compared with a first threshold value; and when the distance exceeds a first threshold value, determining that the circuit breaker trolley is not in the working position, when the distance is lower than the first threshold value, using a second convolutional neural network model to process the latest image to predict a probability value that the circuit breaker trolley is in the working position, and when the probability value exceeds a second threshold value, determining that the circuit breaker trolley is in the working position.

Description

Technical Field

[0001] This disclosure relates to the field of digital technology for electrical equipment, and more specifically, to a method and apparatus for identifying the position of a circuit breaker trolley in a medium-voltage switchgear. Background Technology

[0002] Medium-voltage switchgear is a critical piece of equipment in power distribution systems and a core component ensuring the safe operation of these systems. Maintenance of the switchgear is essential for reliable operation. For maintenance, the circuit breaker trolley must be removed from its working position. After maintenance, the circuit breaker trolley must be returned to its accurate working position; otherwise, power supply will be interrupted. Furthermore, the safety of maintenance personnel must be ensured.

[0003] The position of the circuit breaker trolley must be identified before the switch is closed. High accuracy in identification is also essential for operation.

[0004] Currently, proximity indicators and microswitches are mainly used to identify the position of the circuit breaker trolley.

[0005] • Proximity Status Indicator: Install a proximity status indicator to display the working and testing positions of the circuit breaker trolley. When the circuit breaker trolley moves, it will trigger a status change in the indicator, emitting a corresponding position signal and determining the position of the circuit breaker trolley.

[0006] • Microswitch: A microswitch can be used to detect the position of the circuit breaker trolley. When the circuit breaker trolley reaches the working position or the test position, the microswitch will activate and send position information feedback to the control system. Summary of the Invention

[0007] A brief overview of the invention is given below to provide a basic understanding of certain aspects of it. It should be understood that this overview is not an exhaustive summary of the invention. It is not intended to identify key or essential parts of the invention, nor is it intended to limit the scope of the invention. Its purpose is merely to present certain concepts in a simplified form as a prelude to the more detailed description that follows.

[0008] In a first aspect, a method for identifying the position of a circuit breaker trolley in a medium-voltage switchgear is provided. The method includes acquiring multiple consecutive images of the circuit breaker trolley from a camera monitoring the trolley; calculating the distance between each consecutive image and a reference image indicating that the circuit breaker trolley is in a working position; analyzing the trend of the calculated distances; determining that the circuit breaker trolley is not in a working position when the trend indicates that the circuit breaker trolley is moving; detecting the presence of a quincunx contact in the latest image of the consecutive images using a first convolutional neural network model when the trend indicates that the circuit breaker trolley is stable; determining that the circuit breaker trolley is not in a working position when a quincunx contact is detected; comparing the distance between the latest image and the reference image with a first threshold when the distance exceeds the first threshold; determining that the circuit breaker trolley is not in a working position when the distance is below the first threshold; processing the latest image using a second convolutional neural network model to predict the probability value of the circuit breaker trolley being in a working position when the distance is below the first threshold; and determining that the circuit breaker trolley is in a working position when the probability value exceeds a second threshold, otherwise determining that the circuit breaker trolley is not in a working position.

[0009] This method eliminates the need for additional mechanical sensors, such as proximity indicators and microswitches, reducing system complexity and maintenance costs, while improving reliability by avoiding electromagnetic interference and mechanical wear issues associated with traditional sensing methods.

[0010] In one possible implementation, the method may include analyzing the trend of the calculated distance by determining whether the distance shows an upward trend, a downward trend, or a stable trend, wherein an upward or downward trend indicates that the circuit breaker trolley is moving.

[0011] In this way, trend analysis provides real-time detection of the circuit breaker trolley's movement, enabling immediate identification of non-operating positions without completing a full identification process, thereby improving response time and operational safety.

[0012] Optionally, the method may also include training a first convolutional neural network model on images of plum blossom tentacles collected from different perspectives.

[0013] In this way, training the first neural network model on images from multiple perspectives enhances the robustness and accuracy of plum blossom contact detection under various operating conditions and camera positions.

[0014] Optionally, the method may further include configuring a second convolutional neural network model as a multilayer convolutional neural network trained on images classified as circuit breaker trolley in a working position and circuit breaker trolley in a non-working position.

[0015] In this way, the multi-layered architecture of the second neural network model enables complex feature extraction and pattern recognition, providing accurate probability-based location classification to distinguish between working and non-working locations with high precision.

[0016] Optionally, the method may also include a first threshold predefined based on the test results of the distinguishing image and the reference image.

[0017] By using empirical testing to predefine a first threshold, optimal distinction between acceptable and unacceptable image variations is ensured, providing a reliable filtering mechanism and reducing false positives.

[0018] Optionally, the method may also include defining a second threshold based on the test results to distinguish between working and non-working positions.

[0019] The second threshold, defined through system testing, optimizes the probability-based decision boundary, ensuring accurate classification while minimizing false positives and false negatives.

[0020] Optionally, at least five of the latest images captured by the camera can be acquired as a series of consecutive images.

[0021] Using at least five consecutive images provides sufficient data for reliable trend analysis and statistical validation, enhancing the accuracy of motion detection and location recognition while maintaining computational efficiency.

[0022] In a second aspect, a device for identifying the position of a circuit breaker trolley in a medium-voltage switchgear is provided. The device includes an image acquisition unit configured to acquire multiple consecutive images of the circuit breaker trolley from a camera monitoring the trolley; a calculation unit configured to calculate the corresponding distance between each consecutive image and a reference image indicating that the circuit breaker trolley is in a working position; an analysis unit configured to analyze the trend of the calculated distances; a detection unit configured to use a first convolutional neural network model to detect the presence of a staggered contact in the latest image of the consecutive images when the trend indicates that the circuit breaker trolley is stable; a comparison unit configured to compare the distance between the latest image and the reference image with a first threshold when no staggered contact is detected; a probability prediction unit configured to use a second convolutional neural network model to process the latest image to predict the probability value that the circuit breaker trolley is in a working position when the distance is lower than the first threshold; a first determination unit configured to determine that the circuit breaker trolley is in a working position when the probability value exceeds a second threshold; and a second determination unit configured to determine that the circuit breaker trolley is not in a working position.

[0023] This device architecture provides a modular and scalable circuit breaker trolley position identification solution that can be integrated into existing switchgear control systems without requiring hardware modifications to the switchgear itself, while providing enhanced accuracy through a multi-stage verification process.

[0024] In a third aspect, an electronic device is provided. The electronic device includes: at least one processor; and a memory coupled to said at least one processor, the memory for storing instructions that, when executed by said at least one processor, cause the processor to perform the method according to any one of the first aspects.

[0025] In a fourth aspect, a non-transitory machine-readable storage medium is provided. This non-transitory machine-readable storage medium stores computer instructions, wherein the computer instructions, when executed by a processor, implement the method according to any one of the first aspects.

[0026] In a fifth aspect, a computer program product is provided, which is tangibly stored on a computer-readable medium and includes computer-executable instructions that, when executed, cause at least one processor to perform the method according to any one of the first aspects.

[0027] The method and apparatus according to embodiments of this disclosure have at least one of the following technical advantages:

[0028] This method requires no additional sensor components, uses images from a camera as the data source, and reduces the overall system cost and the cost of further maintenance.

[0029] This method achieves non-contact location identification through image-based analysis, eliminating physical contact between the sensor and the circuit breaker trolley, thus reducing mechanical wear and the need for maintenance intervention. A multi-stage verification process, combining trend analysis, neural network detection, and probabilistic evaluation, provides redundant verification layers, enhancing the reliability of location determination and reducing the risk of false identification due to a single sensor failure.

[0030] This method adapts to different switchgear configurations and circuit breaker trolley designs, eliminating the need for custom hardware sensors for each specific application. Camera positioning and neural network models can be tailored to various switchgear types, providing flexible deployment options and reducing system customization costs.

[0031] This method provides performance enhancement capabilities through continuous learning and improvement of the neural network model. As the system accumulates more operational data and position recognition experience, the neural network model can be updated through additional training to improve accuracy, while traditional mechanical sensors maintain fixed performance characteristics throughout their lifespan. An adaptive threshold adjustment mechanism enables the system to respond to changes in environmental and operational conditions, maintaining consistent recognition accuracy across different operating scenarios. Attached Figure Description

[0032] The above and other objects, features, and advantages of the present invention will be more readily understood by referring to the following description of embodiments of the invention in conjunction with the accompanying drawings. The components in the drawings are merely for illustrating the principles of the invention. In the drawings, the same or similar technical features or components will be represented by the same or similar reference numerals. In the drawings:

[0033] Figure 1 This is a flowchart illustrating an exemplary process of a method for identifying the position of a circuit breaker trolley in a switchgear according to an embodiment of this disclosure;

[0034] Figure 2 A block diagram illustrating an exemplary configuration of a device for identifying the position of a circuit breaker trolley in a medium-voltage switchgear according to another embodiment of this disclosure; and

[0035] Figure 3 A block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0036] The reference numerals in the attached figures are as follows:

[0037] 100: Method for identifying the position of the circuit breaker trolley in a medium-voltage switchgear S102, S104, S106, S108, S110, S112, S114, S116: Steps 200: Identification device for the position of the circuit breaker trolley in medium-voltage switchgear 202: Image Acquisition Unit 204: Computing Unit 206: Analysis Unit 208: Detection Unit 210: Comparison Unit 212: Probability Prediction Unit 214: First Determined Unit 216: Second Determining Unit 300: Computing equipment 302: Processor 304: Memory Detailed Implementation

[0038] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed merely to enable those skilled in the art to better understand and implement the subject matter described herein, and are not intended to limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the elements discussed without departing from the scope of this disclosure. Various processes or components may be omitted, substituted, or added as needed in the various examples.

[0039] In the description of this application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0040] As used herein, the term "comprising" and its variations are open terms meaning "including but not limited to". The term "based on" means "at least partially based on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other definitions, whether explicit or implicit, may be included below. Unless explicitly indicated by the context, the definition of a term shall remain consistent throughout the specification.

[0041] This disclosure provides a method for identifying the position of a circuit breaker trolley in a medium-voltage switchgear. This method uses image data analysis technology combined with a neural network model to determine whether the circuit breaker trolley is in a working position or not. The technical solution according to embodiments of this disclosure is described below with reference to the accompanying drawings.

[0042] Figure 1 This is a flowchart illustrating an exemplary process of a method 100 for identifying the position of a circuit breaker trolley in a switchgear according to an embodiment of this disclosure.

[0043] Method 100 begins with step S102, where multiple consecutive images of the circuit breaker trolley are acquired from the camera monitoring the circuit breaker trolley.

[0044] Specifically, in step S102, sequential image data can be captured at predetermined time intervals to establish a time series for analysis.

[0045] Preferably, the multiple consecutive images include at least five recent images captured by the camera, providing sufficient data points for trend analysis and location determination.

[0046] The camera can be positioned at different locations and angles relative to the circuit breaker trolley within the medium-voltage switchgear to capture images from various viewpoints, thereby improving detection accuracy. The specific camera positioning depends on the switchgear configuration and maintenance requirements, and is not described in detail in this disclosure.

[0047] After acquiring the images, method 100 proceeds to step S104, calculating the corresponding distance between each consecutive image and the reference image representing the circuit breaker trolley in the working position.

[0048] Step S104 applies a distance calculation algorithm to quantify the similarity or difference between each captured image and the reference image. The distance calculation provides a numerical value representing the degree of correspondence between the current circuit breaker trolley position and the reference working position.

[0049] Specifically, step S104 can employ different distance calculation algorithms based on specific application requirements and image features. For example, pixel-level comparison algorithms calculate distances by comparing individual pixel values ​​between images, providing detailed spatial correspondence analysis. Feature-based matching algorithms identify and compare unique features within images, such as edges, corners, or geometric patterns, to determine positional relationships. Histogram comparison algorithms analyze the distribution of pixel intensity or color values ​​to evaluate overall image similarity.

[0050] In step S104, the selection of the distance calculation algorithm can take into account factors such as computational efficiency, accuracy requirements, and robustness to changes in lighting or image noise. Pixel-level comparison offers high accuracy but requires significant computational resources. Feature-based matching offers good accuracy with moderate computational requirements. Histogram comparison provides fast processing and acceptable accuracy for many applications.

[0051] In a specific example, distance values ​​D1, D2, D3, D4, and D5 corresponding to consecutive images can be generated, creating a numerical measurement sequence representing the time-varying position of the circuit breaker trolley relative to a reference image. These distance values ​​form the basis for subsequent trend analysis and position determination steps in method 100.

[0052] Next, method 100 proceeds to step S106, which analyzes the trends of the various distances calculated in the above steps to determine whether the circuit breaker trolley is moving or in a stable state.

[0053] For example, step S106 processes the distance values ​​D1, D2, D3, D4, and D5 generated in step S104 to identify states indicating whether the circuit breaker trolley is moving or stable. The trend analysis in step S106 provides a basis for determining whether further position verification steps should be performed or whether the conclusion that the circuit breaker trolley is not in the working position can be drawn immediately.

[0054] Step S106 analyzes whether the calculated distances show an upward, downward, or stable trend. An upward trend indicates the circuit breaker trolley is moving away from its working position when the distance values ​​gradually increase in a continuous image. A downward trend indicates the circuit breaker trolley is moving towards its working position when the distance values ​​gradually decrease. A stable trend indicates the circuit breaker trolley is stationary when the distance values ​​remain relatively constant.

[0055] If the trend in step S106 is an upward or downward trend, it indicates that the circuit breaker trolley is moving. Method 100 proceeds to step S116, determines that the circuit breaker trolley is not in the working position, and ends without performing any other processing steps. This is because if the circuit breaker trolley is moving, it means that the circuit breaker trolley is necessarily not positioned in the working position.

[0056] In an alternative implementation, an output signal or status indication that the circuit breaker trolley is not in the working position can be output when movement is detected.

[0057] When the trend analysis in step S106 indicates that the circuit breaker trolley is stable, method 100 continues to step S108, using the first convolutional neural network model to detect whether there is a quincunx contact in the latest image of the continuous images.

[0058] Step S108 is the second verification stage in the multi-stage identification process. If a plum blossom contact is present in the latest captured image, it indicates that the circuit breaker trolley is not positioned in the working position.

[0059] The plum blossom contact represents the conductive component of the circuit breaker trolley, becoming visible when the trolley is not fully inserted into the operating position. When the circuit breaker trolley is correctly positioned in the operating position, the plum blossom contact remains hidden or is obscured within the switchgear housing. Therefore, the visibility of the plum blossom contact in a captured image can be used as an indicator that the circuit breaker trolley has not reached the operating position, even if the circuit breaker trolley appears stable based on trend analysis.

[0060] Step S108 applies a first convolutional neural network model specifically trained for plum blossom contact detection. The first convolutional neural network model processes the latest image in a series of images to identify the presence or absence of plum blossom contact features within the image. The first convolutional neural network model outputs a detection result indicating whether a plum blossom contact exists in the analyzed image.

[0061] In one exemplary implementation, the first convolutional neural network model is trained using images of the cloverleaf contact from different perspectives to ensure robust detection performance across various camera positions and orientations. For example, training images may include cloverleaf contacts captured from above, from the side, at different distances, and under varying lighting conditions. This diverse training dataset enables the first convolutional neural network model to recognize cloverleaf contact features regardless of the specific camera positioning or environmental conditions during operation. It is understood that, in addition to the convolutional neural network model, other types of neural network models can be trained as needed; the specific training process for neural network models is not described in detail herein.

[0062] When step S108 detects the presence of a cloverleaf contact in the latest image, method 100 proceeds to step S116, determining that the circuit breaker trolley is not in the operating position. The presence of the cloverleaf contact is clear evidence that the circuit breaker trolley has not reached the operating position, regardless of the stability of the trend analysis indication.

[0063] If step S108 does not detect the presence of the quincunx contact in the latest image, method 100 proceeds to step S110, which compares the distance between the latest image and the reference image with a first threshold. Step S110 represents the third verification stage in the multi-stage identification process, wherein method 100 evaluates the distance value between the latest image and the reference image to determine whether the latest image sufficiently corresponds to the reference image representing that the circuit breaker trolley is in the working position.

[0064] Step S110 processes the distance value D5 between the latest image and the reference image calculated in step S104. The distance value D5 represents a quantitative measurement of the similarity or difference between the current circuit breaker trolley position captured in the latest image and the reference working position represented by the reference image. Step S110 compares the distance value D5 with a predetermined first threshold to determine whether the circuit breaker trolley position is within the acceptable tolerance range of the working position.

[0065] Specifically, the first threshold can be predefined based on test results distinguishing the image from the reference image. The test results are obtained by systematically evaluating the distance values ​​calculated between the known working position image and the reference image, as well as the distance values ​​calculated between the known non-working position image and the reference image. The test results establish a statistical distribution of the distance values, enabling the determination of an appropriate threshold that reliably distinguishes between working and non-working position states.

[0066] It is understandable that the first threshold can be dynamically adjusted based on environmental conditions, lighting changes, or historical performance data, rather than remaining fixed throughout the operation.

[0067] When it is determined in step S110 that the distance value D5 exceeds the first threshold, it indicates that the similarity between the latest image and the reference image is low, that is, the position of the circuit breaker trolley deviates from the reference working position beyond the acceptable tolerance. Method 100 proceeds to step S116 to determine that the circuit breaker trolley is not in the working position.

[0068] When it is determined in step S110 that the distance value D5 is lower than the first threshold, method 100 continues to step S112, using the second convolutional neural network model to process the latest image to predict the probability value that the circuit breaker trolley is in the working position.

[0069] Step S112 represents the fourth verification stage in the multi-stage identification process, where method 100 applies machine learning techniques to generate a probability assessment of the circuit breaker trolley position based on image analysis.

[0070] Step S112 processes the latest image from the successive images captured during step S102 using a second convolutional neural network model specifically designed for circuit breaker cart position classification. The second convolutional neural network model analyzes visual features within the latest image to determine the probability that the circuit breaker cart is located in a working or non-working position. For example, step S112 may generate probability values ​​representing the confidence level of the circuit breaker cart being in a working position, providing a numerical measurement between 0 and 1, where values ​​closer to 1 indicate a higher confidence level that the circuit breaker cart is in a working position.

[0071] The second convolutional neural network model consists of a multi-layer convolutional neural network trained using images classified as either the circuit breaker trolley in a working position or in a non-working position.

[0072] Classifying training images into working and non-working position categories provides the foundation for the supervised learning process. Working position images represent circuit breaker trolleys correctly positioned and fixed in the working position, ready for switching operations. Non-working position images include circuit breaker trolleys in test positions, intermediate positions between working and test positions, or any other position that does not meet the safety requirements for switching operations. The binary classification method enables the second convolutional neural network model to specifically focus on the distinction between acceptable and unacceptable circuit breaker trolley positions.

[0073] Following the probability prediction in step S112, method 100 proceeds to step S114, whereby the circuit breaker trolley is determined to be in the working position when the probability value exceeds a second threshold. Step S114 represents the final verification stage in the multi-stage identification process, where method 100 compares the predicted probability value with a predetermined threshold to make a deterministic position determination. When the probability value indicates a sufficiently high confidence that the circuit breaker trolley is positioned in the working position, step S114 determines that the circuit breaker trolley is in the working position.

[0074] The second threshold can be predefined based on the test results that distinguish between working and non-working positions.

[0075] When the probability value in step S114 exceeds the second threshold, method 100 determines that the circuit breaker trolley is in the working position. The determination in step S114 represents the completion of a multi-stage verification process, confirming that the circuit breaker trolley has passed all verification stages, including trend stability, invisible clover contact, distance threshold satisfaction, and probability threshold satisfaction. After step S114, an output signal indicating that the circuit breaker trolley is in the working position can be further generated, ensuring that switching operations can be performed when the circuit breaker trolley is correctly positioned and can be operated safely.

[0076] If the probability value in step S114 does not exceed the second threshold, method 100 proceeds to step S116, determining that the circuit breaker trolley is not in the working position. This determination in step S116, based on insufficient probability, indicates that the second convolutional neural network model lacks sufficient confidence in classifying the working position, even though the circuit breaker trolley passed the previous verification phase. Step S116 generates an output indicating that the circuit breaker trolley's position remains uncertain, and the switching operation should not be performed until correct positioning is confirmed through additional verification or manual inspection.

[0077] Figure 2 This is a block diagram of an exemplary configuration of a medium-voltage switchgear circuit breaker trolley position identification device 200 according to another embodiment of this disclosure.

[0078] The device 200 includes an image acquisition unit 202, which is configured to acquire multiple consecutive images of the circuit breaker trolley from a camera monitoring the circuit breaker trolley.

[0079] Image acquisition unit 202 is connected to calculation unit 204, which is configured to calculate the corresponding distance between each consecutive image and a reference image representing the circuit breaker trolley in the working position.

[0080] The calculation unit 204 is connected to the analysis unit 206, which is configured to analyze the trend of the calculated distance. The analysis unit 206 can implement statistical analysis algorithms to determine whether the distance sequence exhibits a stable, increasing, or decreasing trend.

[0081] Analysis unit 206 is connected to second determination unit 216 and detection unit 208. Second determination unit 216 is configured to determine that the circuit breaker trolley is not in the operating position and receives input from multiple units within device 200. When analysis unit 206 determines a trend indicating that the circuit breaker trolley is moving, it provides output directly to second determination unit 216.

[0082] The detection unit 208 is configured to use a first convolutional neural network model to detect whether a quincunx contact exists in the latest image of a series of images when the trend indicates that the circuit breaker trolley is stable.

[0083] The detection unit 208 may include neural network processing hardware, including a dedicated inference engine or tensor processing unit, to efficiently execute the first convolutional neural network model. The detection unit 208 may store the trained model parameters in a dedicated memory and implement an inference algorithm optimized for real-time plum blossom contact detection.

[0084] The detection unit 208 is connected to the second determination unit 216 and the comparison unit 210. When a clover contact is detected, the detection unit 208 provides an output to the second determination unit 216, indicating that the circuit breaker trolley is not in the working position.

[0085] The comparison unit 210 is configured to compare the distance between the latest image and the reference image with a first threshold when no quincunx contact is detected.

[0086] The comparison unit 210 can implement threshold comparison logic and store configurable thresholds that can be adjusted based on system calibration and operational experience. The comparison unit 210 is connected to the second determination unit 216 and the probability prediction unit 212.

[0087] The probability prediction unit 212 is configured to process the latest image using a second convolutional neural network model to predict the probability that the circuit breaker trolley is in the working position when the distance between the latest image and the reference image is less than a first threshold. The probability prediction unit 212 includes advanced neural network processing capabilities, including a floating-point unit for probability calculation. This unit stores the trained second convolutional neural network model and implements an inference algorithm optimized for the probability prediction task.

[0088] The probability prediction unit 212 is connected to the first determination unit 214 and the second determination unit 216. The first determination unit 214 is configured to determine that the circuit breaker trolley is in the working position when the probability value exceeds a second threshold. The first determination unit 214 implements the final decision logic for determining the position of the circuit breaker trolley.

[0089] Details of the operation and function of the various parts of the medium-voltage switchgear circuit breaker trolley position identification device 200 according to embodiments of this disclosure can be found, for example, in conjunction with reference to [reference needed]. Figure 1 The relevant parts of the embodiments of the method for identifying the position of the circuit breaker trolley in a medium-voltage switchgear according to the present disclosure are the same or similar, and will not be described in detail here.

[0090] It should be noted that, Figure 2 The structure of the identification device 200 for the position of the circuit breaker trolley in the medium-voltage switchgear shown, and its constituent units, is merely exemplary. Those skilled in the art can modify it as needed. Figure 2 The structural diagram shown is modified.

[0091] As per the above reference Figures 1 to 2 The methods and apparatus according to embodiments of this disclosure have been described. Each unit of the medium-voltage switchgear circuit breaker trolley position identification device 200 described above can be implemented in hardware, software, or a combination of hardware and software.

[0092] Figure 3 A block diagram of an electronic device 300 according to an embodiment of the present disclosure is shown. According to one embodiment, the electronic device 300 may include at least one processor 302 that executes at least one computer-readable instruction stored or encoded in a computer-readable storage medium (i.e., memory 304).

[0093] It should be understood that the computer-executable instructions stored in memory 304, when executed, cause at least one processor 302 to perform the above-described combinations in the various embodiments of this disclosure. Figure 1-2 The description includes various operations and functions.

[0094] According to one embodiment, a non-transitory machine-readable medium is provided. This non-transitory machine-readable medium may have machine-executable instructions (i.e., the elements implemented in software as described above), which, when executed by a machine, cause the machine to perform the above-described combinations in the various embodiments of this disclosure. Figure 1-2 The description includes various operations and functions.

[0095] According to one embodiment, a computer program is provided, including computer-executable instructions that, when executed, cause at least one processor to perform the above-described embodiments of the present disclosure. Figure 1-2 The description includes various operations and functions.

[0096] According to one embodiment, a computer program product is provided, including computer-executable instructions that, when executed, cause at least one processor to perform the above-described embodiments of the present disclosure. Figure 1-2 The description includes various operations and functions.

[0097] Not all units in the above structural diagrams are necessary; some units can be omitted as needed. The device structures described in the above embodiments can be physical structures or logical structures. That is, some units may be implemented by the same physical entity, or some units may be implemented by multiple physical entities, or they may be jointly implemented by certain components in multiple independent devices.

[0098] The foregoing description of this disclosure is provided to enable any person skilled in the art to implement or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein can be applied to other variations without departing from the scope of this disclosure. Therefore, this disclosure is not limited to the examples and designs described herein, but is consistent with the widest scope of the principles and novel features disclosed herein.

[0099] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0100] In this patent application, nouns and pronouns relating to people are not limited to specific genders.

Claims

1. A method (100) for identifying the position of a circuit breaker trolley in a medium-voltage switchgear, comprising: (S102) Obtain multiple consecutive images of the circuit breaker trolley from the camera monitoring the circuit breaker trolley; Calculate (S104) the corresponding distance between each consecutive image and the reference image representing the circuit breaker trolley in the working position; Analyze the trends of each distance calculated in (S106); When the trend indicates that the circuit breaker trolley is moving, determine (S116) that the circuit breaker trolley is not in the working position; When the trend indicates that the circuit breaker trolley is stable, the first convolutional neural network model is used to detect (S108) whether there is a quincunx contact in the latest image among multiple consecutive images; When the plum blossom contact is detected, it is determined (S116) that the circuit breaker trolley is not in the working position; When the plum blossom contact is not detected, the distance between the latest image and the reference image is compared with a first threshold (S110). When the distance between the latest image and the reference image exceeds the first threshold, it is determined (S116) that the circuit breaker trolley is not in the working position; When the distance is lower than the first threshold, the latest image is processed using a second convolutional neural network model (S112) to predict the probability value that the circuit breaker trolley is in the working position; as well as When the probability value exceeds the second threshold, it is determined (S114) that the circuit breaker trolley is in the working position; otherwise, it is determined (S116) that the circuit breaker trolley is not in the working position.

2. The method (100) according to claim 1, wherein, The analysis of the calculated trends of each distance includes determining whether the distance shows an upward trend, a downward trend, or a stable trend, wherein the upward trend or the downward trend indicates that the circuit breaker trolley is moving.

3. The method (100) according to claim 1 or 2, wherein, The first convolutional neural network model was trained using images of plum blossom tentacles collected from different perspectives.

4. The method (100) according to claim 1 or 2, wherein, The second convolutional neural network model includes a multi-layer convolutional neural network trained using images classified as circuit breaker trolley in working position and circuit breaker trolley in non-working position.

5. The method (100) according to claim 1 or 2, wherein, The first threshold is predefined based on the test results that distinguish the image from the reference image.

6. The method (100) according to claim 1 or 2, wherein, The second threshold is predefined based on the test results that distinguish between working position and non-working position states.

7. The method (100) according to claim 1 or 2, wherein, The multiple consecutive images include at least five recent images captured by the camera.

8. A device (200) for identifying the position of a circuit breaker trolley in a medium-voltage switchgear, comprising: The image acquisition unit (202) is configured to acquire multiple consecutive images of the circuit breaker trolley from the camera monitoring the circuit breaker trolley; The calculation unit (204) is configured to calculate the corresponding distance between each consecutive image and a reference image representing the circuit breaker trolley in the working position; The analysis unit (206) is configured to analyze the trend of distance calculation; The detection unit (208) is configured to use a first convolutional neural network model to detect whether a quincunx contact exists in the latest image of a series of images when the trend indicates that the circuit breaker trolley is stable; The comparison unit (210) is configured to compare the distance between the latest image and the reference image with a first threshold when no quincunx contact is detected; The probability prediction unit (212) is configured to process the latest image using a second convolutional neural network model to predict the probability value of the circuit breaker trolley being in the working position when the distance between the latest image and the reference image is less than the first threshold. as well as The first determining unit (214) is configured to determine that the circuit breaker trolley is in the working position when the probability value exceeds the second threshold. The second determining unit (216) is configured to determine that the circuit breaker trolley is not in the working position based on information from the analysis unit (206), the detection unit (208), the comparison unit (210), and the probability prediction unit (212).

9. Electronic equipment (300), including: At least one processor (302); as well as A memory (304) coupled to the at least one processor (302), the memory being used to store instructions that, when executed by the at least one processor (302), cause the processor (302) to perform the method as described in any one of claims 1 to 7.

10. A non-transitory machine-readable storage medium storing computer instructions that, when executed, cause a machine to perform the method as described in any one of claims 1-7.

11. A computer program product tangibly stored on a computer-readable medium and comprising computer-executable instructions that, when executed, cause at least one processor to perform the method according to any one of claims 1-7.

Images