Remote intelligent circuit breaker opening and closing detection system
The remote intelligent circuit breaker opening and closing detection system utilizes deep learning and Faster-RCNN object detection network to solve the problems of subjective misjudgment and low recognition accuracy in circuit breaker equipment status identification, thus achieving efficient and safe circuit breaker status identification and monitoring.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-20
- Publication Date
- 2026-03-03
AI Technical Summary
Existing technologies rely on manual verification for circuit breaker equipment status identification, which carries the risk of subjective misjudgment and is inefficient. Machine learning algorithms have low accuracy in complex environments, making it difficult to achieve full-process automation of the power system.
The system employs a remote intelligent circuit breaker opening and closing detection system. It utilizes deep learning technology and the Faster-RCNN target detection network, combined with image sensors and edge computing boxes, to achieve real-time identification and data feedback of the circuit breaker's opening and closing status. It is also equipped with a linkage supplementary lighting device to adapt to complex environments.
It improves the automation level and detection efficiency of circuit breaker status identification, reduces the labor intensity and safety risks of operators, enhances the accuracy of identification and the compatibility of equipment, and forms a complete remote monitoring closed loop.
Smart Images
Figure CN223966670U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of circuit breaker detection technology, specifically to a remote intelligent circuit breaker opening and closing detection system. Background Technology
[0002] With the continuous advancement of smart grid construction, significant progress has been made in the remote transformation of power system switching operations. Traditional manual on-site operation modes are gradually being replaced by remote centralized control systems. However, in actual operation processes, the crucial step of manually verifying the status of electrical equipment on-site still requires significant effort. Since the mechanical condition of primary equipment such as high-voltage circuit breakers and disconnectors directly affects the operational safety of the power system, accurate identification of their position status has become a vital guarantee for ensuring the correctness of switching operations.
[0003] In existing technologies, equipment status verification mainly relies on manual visual inspection. This method faces several technical obstacles in practical application: First, there are significant differences in the design of on / off status indicators for high-voltage equipment from different manufacturers and models. Specifically, there is a lack of standardized elements such as font specifications, color schemes (e.g., contrasting red and green or single-color markings), and shape design (circular indicator windows versus arrow-shaped signs). Second, outdoor equipment is exposed to complex environments for extended periods, leading to issues such as paint peeling, color fading, and blurred text on the signs, especially under adverse weather conditions like rain, snow, fog, and haze, further reducing legibility. More importantly, manual inspection carries the risk of subjective judgment bias; inexperienced operators are prone to misjudging the status, and the safety hazards associated with close contact with high-voltage equipment cannot be eliminated.
[0004] To address the aforementioned issues, existing technologies have attempted to employ intelligent detection schemes based on machine learning, but the practical application results remain unsatisfactory. Traditional algorithms, such as Support Vector Machines and Random Forests, rely on manually designed feature templates. However, the environment at power plant sites is characterized by a wide variety of equipment, variable installation angles, and unstable lighting conditions, making manually constructed feature models difficult to adapt to the diverse detection scenarios. Particularly under interference conditions such as equipment surface reflections, component occlusion, and visual distortion, the stability of feature extraction by traditional algorithms significantly decreases, leading to persistently high false positive and false negative rates.
[0005] The deficiencies in the current technological system have had a substantial impact on the safe operation of the power system. Manual verification is inefficient and unreliable, while intelligent detection equipment faces dual bottlenecks in environmental adaptability and detection accuracy. This makes it difficult to truly automate the entire switching operation process. Especially at critical nodes such as hub substations and ultra-high-voltage transmission lines, accurate identification of equipment status has become a significant technological shortcoming restricting the construction of smart grids, urgently requiring technological improvements to build a safer and more reliable detection system. Utility Model Content
[0006] The purpose of this invention is to address the limitations of existing technologies in circuit breaker equipment status identification, such as the significant influence of subjective human factors, high workload, low detection accuracy, and difficulties in implementing machine recognition algorithms. Therefore, a remote intelligent circuit breaker opening and closing detection system is proposed. This system utilizes existing deep learning technology to realize the hardware connection and auxiliary component deployment scheme of each component module, becoming an easy-to-use and efficient system for remote real-time detection and data feedback of circuit breaker opening and closing status, providing technical support for the construction of intelligent power plants.
[0007] The present invention employs the following technical solution to achieve its objective:
[0008] A remote intelligent circuit breaker opening and closing detection system is disclosed. The system consists of a status monitoring subsystem and a remote service subsystem connected by communication. The status monitoring subsystem includes at least one detection terminal, each of which has an image sensor and an edge computing box connected by communication. The installation position of the image sensor is matched with the area where the circuit breaker is located, so that the image sensor can acquire image data corresponding to the circuit breaker opening and closing components and status indicators. The edge computing box is used to output the circuit breaker opening and closing information to the remote service subsystem after edge computing.
[0009] Furthermore, the edge computing box is equipped with an NPU unit, which deploys a Faster-RCNN-based object detection network. This object detection network outputs circuit breaker opening and closing information after edge computing.
[0010] Specifically, the image sensor establishes a video stream communication connection with the edge computing box via the RTSP protocol; the edge computing box is also equipped with a video encoding unit, which is used to re-encode the image data and push the video stream transmitted via the RTSP protocol to the remote service subsystem.
[0011] Preferably, each detection terminal also has a linkage supplementary lighting device, the illumination area of which matches the area where the circuit breaker is located, and when the linkage supplementary lighting device emits light, the circuit breaker's opening and closing components and status indicators are illuminated; the linkage supplementary lighting device is communicatively connected to the edge computing box.
[0012] Specifically, the linkage supplementary lighting device includes a photoresistor, a supplementary light, and an ADC module. The photoresistor and the supplementary light are both electrically connected to the ADC module, and the ADC module is communicatively connected to the edge computing box. The photoresistor is used to detect the ambient light intensity in the area where the circuit breaker is located, and the supplementary light is used to emit light when a supplementary light signal is received.
[0013] Specifically, the ADC module is connected to the serial communication unit of the edge computing box. The ADC module has two computing modes: interrupt mode and polling mode. In interrupt mode, the edge computing box switches to ambient light processing mode by triggering a hardware interrupt, and at the same time sends a supplementary light signal to the supplementary light. In polling mode, the edge computing box receives polling requests from the edge computing box.
[0014] Furthermore, the remote service subsystem includes a multi-layer switch, an NVR server, a detection server, and a remote monitoring platform. The multi-layer switch connects to the edge computing box, the NVR server, and the detection server, respectively, and the detection server communicates with the remote monitoring platform.
[0015] Specifically, the multilayer switch is used to provide PoE power support for the image sensors and edge computing boxes in the detection terminal, while providing Ethernet network switching and VLAN isolation support for the edge computing boxes, NVR servers, and detection servers.
[0016] Specifically, the communication connection protocols between the multilayer switch and the edge computing box include the RTSP protocol and the WebSocket protocol. The RTSP protocol is used to transmit video streams from the image sensor transmitted through the edge computing box, and the WebSocket protocol is used to transmit circuit breaker opening and closing information from the edge computing box. The communication connection protocol between the multilayer switch and the NVR server is the RTSP protocol. The NVR server is used to provide video stream segment storage and video stream relay support between various components.
[0017] Preferably, the detection server is used to feed back the circuit breaker opening and closing information obtained by the multilayer switch to the remote monitoring platform in real time; the communication connection between the detection server and the remote monitoring platform adopts L2TP encrypted transmission connection, and the communication connection protocol used is WebSocket protocol.
[0018] In summary, due to the adoption of this technical solution, the beneficial effects of this utility model are as follows:
[0019] The implementation of this invention significantly improves the automation level and detection efficiency of circuit breaker status identification. By deploying intelligent detection terminals and remote monitoring platforms with deep learning capabilities, the system can collect and analyze multi-dimensional feature information of the circuit breaker's open and closed status in real time, achieving unattended operation around the clock. During the detection process, there is no need for manual close contact with high-voltage equipment, greatly reducing the labor intensity and safety risks for operators. At the same time, the digital transmission channel ensures the timeliness and integrity of the status data, providing a reliable basis for power grid dispatching decisions.
[0020] The system's architecture design effectively improves recognition accuracy under complex operating conditions. The integrated high-precision image sensor and linked supplementary lighting device maintain relatively stable detection performance under interference conditions such as blurred markings on the circuit breaker surface and sudden changes in ambient light. Compared to traditional machine learning recognition methods that rely on manually designed features, this invention applies mature deep learning technology to identify the differentiated markings of different circuit breaker models, thereby significantly enhancing equipment compatibility and reducing dependence on specific marking specifications.
[0021] After implementing this system, the acquisition and processing of circuit breaker equipment status data can be directly integrated into the existing power automation system, forming a complete remote monitoring closed loop. The standardized status information output by the system can be seamlessly connected with the power plant's intelligent management platform, providing data support for intelligent applications such as equipment lifecycle management and fault early warning analysis. In addition, the entire system's architecture is modularly designed, giving it good environmental adaptability and enabling rapid deployment in substations with different voltage levels and installation environments.
[0022] This invention effectively avoids the risk of visual misjudgment inherent in traditional manual inspections, and its non-contact detection mechanism fundamentally eliminates the safety hazards of close-range operation. The remote monitoring platform, based on existing mature technology, can automatically archive received detection data and generate visual reports, thereby improving the traceability of circuit breaker equipment status records and providing a data foundation for preventative maintenance of related power equipment. Compared to existing detection devices or methods, this invention maintains high detection accuracy while reducing daily maintenance costs, demonstrating significant engineering practical value. Attached Figure Description
[0023] Figure 1 This is a schematic diagram of the system structure and connection relationship of this utility model;
[0024] Figure 2 This is a schematic diagram of an example target detection network in this utility model;
[0025] Figure 3 This is a schematic diagram of the linkage supplementary lighting device in this utility model. Detailed Implementation
[0026] To make the objectives, technical solutions, and advantages of the embodiments of this utility model clearer, the technical solutions of the embodiments of this utility model will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this utility model, not all embodiments. The components of the embodiments of this utility model described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.
[0027] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0028] Example 1
[0029] A remote intelligent circuit breaker opening and closing detection system, Figure 1 The diagram illustrates the connection relationships of the key components of the system, which can be viewed simultaneously. This embodiment describes these connections as follows:
[0030] The system consists of a status monitoring subsystem and a remote service subsystem with communication connections. The status monitoring subsystem includes at least one detection terminal, each of which has an image sensor and an edge computing box with communication connections. The installation location of the image sensor is matched with the area where the circuit breaker is located, so that the image sensor can acquire image data corresponding to the circuit breaker's opening and closing components and status indicators. The edge computing box is used to output the circuit breaker's opening and closing information to the remote service subsystem after edge computing.
[0031] In this embodiment, taking a circuit breaker opening / closing detection area as an example, an image sensor is deployed in a location with good visibility in the area where the circuit breaker is located. This image sensor is implemented using an industrial camera, specifically the DH-IPC-HDW1230C-A model. The industrial camera establishes a video stream connection with the edge computing box via the RTSP protocol, thereby enabling real-time transmission of image data corresponding to the circuit breaker opening / closing components and status indicators to the edge computing box for processing.
[0032] The edge computing box contains an NPU unit, which deploys a Faster-RCNN-based object detection network. This object detection network outputs circuit breaker opening and closing information after edge computing. The object detection network in this embodiment is as follows: Figure 2 As shown, in practical implementation, existing mature Faster-RCNN technologies can be directly used to construct the convolutional network feature extractor, attention-based global context extraction module, and multi-scale local context extraction module shown in the figure, thereby performing feature extraction and detection / recognition processing on the input image. After edge computing is performed in the NPU unit within the edge computing box, the circuit breaker opening / closing information can be output, including the current opening / closing status of the circuit breaker, change history, etc., which can then be subsequently utilized by the remote service subsystem.
[0033] The target detection network structure presented in this embodiment is mature and features lightweight design, low latency, and high accuracy. It can detect the circuit breaker's open / closed status in real time, making it suitable for this identification scenario. Furthermore, by configuring the edge computing box logically, it can add alarm information to the circuit breaker's open / closed status information when a change in the circuit breaker's open / closed status is detected, thus alerting relevant operators to take action.
[0034] For edge computing box communication connection transmission feedback, this embodiment uses the WebSocket communication protocol to send data frames of circuit breaker changes to the remote service subsystem, as shown in the following example:
[0035] {
[0036] "timestamp":"2025-03-1T14:30:00Z",
[0037] "location":"Equipment A-Pipeline B",
[0038] "alert_change":1,
[0039] "video_url":"http: / / nvr01.local / video / alert_21.mp4"
[0040] }
[0041] This data frame includes the circuit breaker change time, the circuit breaker change location, and the corresponding circuit breaker change video. Furthermore, the edge computing box can integrate mature image quality assessment algorithms to detect image illumination, image contrast, and other factors in the real-time acquired circuit breaker change video. If image quality issues are found, additional auxiliary components can be called to optimize image acquisition, thereby improving imaging quality and reducing the false alarm rate.
[0042] In this embodiment, one edge computing box connects to one or more industrial cameras, but one edge computing box is only responsible for detecting the opening and closing of a circuit breaker in the area where the circuit breaker is located. The applicable edge computing box model in this scenario is the EC01 model using RK3399+RK1808. This model is an AI edge computing solution based on the Rockchip RK3399 and RK1808 architecture. The RK3399 is a high-performance processor, while the RK1808 is the NPU unit in this embodiment, suitable for structured video analysis of multiple cameras and network cameras. Deploying a target detection network in it to achieve video image detection can be directly implemented by those skilled in the art based on existing mature technologies.
[0043] In this embodiment, after establishing a video stream communication connection based on the RTSP protocol between the industrial camera and the edge computing box, the edge computing box re-encodes the image data using its built-in video encoding unit and pushes the RTSP-transmitted video stream to the remote service subsystem. This video stream corresponds to the data frame under the aforementioned WebSocket communication protocol and may include a unique ID of the current video segment, the time and location of the circuit breaker's open / closed state change, and other information.
[0044] Example 2
[0045] Based on Example 1, this example further optimizes the remote intelligent circuit breaker opening and closing detection system. Each detection terminal in the status monitoring subsystem is also equipped with a corresponding linkage supplementary lighting device. The illumination area of this linkage supplementary lighting device matches the area where the circuit breaker is located. When the linkage supplementary lighting device emits light, the circuit breaker opening and closing components and status indicators are illuminated; the linkage supplementary lighting device is communicatively connected to the edge computing box.
[0046] like Figure 3 As shown, the linkage supplementary lighting device includes a photoresistor, a supplementary light, and an ADC module. The photoresistor and the supplementary light are both electrically connected to the ADC module, and the ADC module is communicatively connected to the edge computing box. The photoresistor is used to detect the ambient light intensity of the area where the circuit breaker is located, and the supplementary light is used to emit light when a supplementary light signal is received.
[0047] In this embodiment, the photoresistor is a GL528, and the ADC module is constructed using an AD7715 analog-to-digital converter. For the linked supplementary lighting device, its important functions include monitoring ambient light and providing supplementary lighting for the circuit breaker. The corresponding signal data generated by the photoresistor based on changes in ambient light is input to the edge computing box after passing through the ADC module. The edge computing box can process the ambient light data, generate a supplementary lighting signal, and then transmit it back to the supplementary lighting lamp to achieve related control.
[0048] The AD7715 analog-to-digital converter connects to the serial communication unit of the edge computing box and has two computing modes: interrupt mode, which triggers a hardware interrupt to switch the edge computing box to ambient light processing mode and simultaneously sends a supplementary light signal to the supplementary light lamp; and polling mode, which receives polling requests from the edge computing box. The information obtained from these polling requests, such as ambient light intensity and supplementary light intensity, can be used to adjust the relevant parameters of the target detection network in the edge computing box, providing assistance for the detection of the circuit breaker's open / closed status.
[0049] Example 3
[0050] Based on any of the above embodiments, this embodiment provides a detailed description of the remote service subsystem in the remote intelligent circuit breaker opening and closing detection system. Its structure and connection relationships can also be found in [reference needed]. Figure 1This is a schematic diagram. The remote service subsystem includes a multi-layer switch, an NVR server, a detection server, and a remote monitoring platform. The multi-layer switch connects to the edge computing box, the NVR server, and the detection server, respectively. The detection server is communicatively connected to the remote monitoring platform.
[0051] In this embodiment, the multilayer switch is model S1750-24G-HPWR, which, together with Cat 5e cable, provides 12V PoE power to the edge computing box and also provides power support for the industrial cameras. Furthermore, the multilayer switch can perform VLAN segmentation based on the location of different circuit breakers, enabling VLAN isolation support for the edge computing box, NVR server, and detection server, ensuring the security of data transmission networks. The communication connection protocols between the multilayer switch and the edge computing box include RTSP and WebSocket protocols. The RTSP protocol is used to transmit video streams from the image sensor transmitted via the edge computing box, and the WebSocket protocol is used to transmit circuit breaker opening and closing information from the edge computing box. Therefore, the presence of the multilayer switch enables the edge computing box, NVR server, and detection server to all have standard Ethernet data exchange capabilities, thus ensuring the network transmission of the entire system.
[0052] The NVR server is model Dahua DH-NVR624-128-4KS2. The NVR server can receive video of circuit breaker opening / closing status changes sent from the edge computing box via a multi-layer switch. Its communication connection protocol with the multi-layer switch also uses the RTSP protocol. Therefore, RTSP video streams from industrial cameras are stored in the NVR server after being transmitted through the multi-layer switch. The main function of the NVR server is to provide segment storage of video streams and support video stream relay between various components.
[0053] In this embodiment, the detection server is used to feed back the circuit breaker opening and closing information obtained by the multilayer switch to the remote monitoring platform in real time via the WebSocket protocol. After receiving the circuit breaker opening and closing information from the edge computing box through the multilayer switch, it parses the data frames of the circuit breaker status changes and can feed back information such as the time of the circuit breaker change, the current opening and closing status, and the location of the circuit breaker to the remote monitoring platform. The communication connection between the detection server and the remote monitoring platform can preferably adopt an L2TP encrypted transmission connection to ensure the accessibility and security of remote monitoring in the industrial field and prevent data from being tampered with or stolen.
[0054] After being relayed through a multi-layer switch, the detection server can retrieve the corresponding circuit breaker status change video from the NVR server using the WebDAV protocol. Thanks to the application of the target detection network in the edge computing box, labeled result boxes can be added to the video information and displayed on the front-end UI of the remote monitoring platform, enabling operators to have a good grasp of the current circuit breaker status.
[0055] The circuit breaker change data frames continuously recorded during the circuit breaker opening / closing status detection process can be stored in the database of the remote monitoring platform. This facilitates the analysis of trends before and after relevant events, optimizing subsequent detection strategies and preventing false or missed detections. By parsing these circuit breaker change data frames and combining them with the video stream data from the NVR server, the detection server enables the remote monitoring platform to perform secondary analysis and supplementary verification, obtaining complete circuit breaker opening / closing information. This allows for efficient and accurate circuit breaker opening / closing status detection and response.
Claims
1. A remote intelligent circuit breaker opening and closing detection system, characterized in that: The system consists of a status monitoring subsystem and a remote service subsystem with communication connections. The status monitoring subsystem includes at least one detection terminal, each of which has an image sensor and an edge computing box with communication connections. The installation location of the image sensor is matched with the area where the circuit breaker is located, so that the image sensor can acquire image data corresponding to the circuit breaker's opening and closing components and status indicators. The edge computing box is used to output the circuit breaker's opening and closing information to the remote service subsystem after edge computing.
2. The remote intelligent circuit breaker opening and closing detection system according to claim 1, characterized in that: The edge computing box is equipped with an NPU unit, which deploys a Faster-RCNN-based object detection network. The object detection network outputs circuit breaker opening and closing information after edge computing.
3. The remote intelligent circuit breaker opening and closing detection system according to claim 2, characterized in that: The image sensor establishes a video stream communication connection with the edge computing box via the RTSP protocol. The edge computing box is also equipped with a video encoding unit, which is used to re-encode the image data and push the video stream transmitted via the RTSP protocol to the remote service subsystem.
4. The remote intelligent circuit breaker opening and closing detection system according to claim 1, characterized in that: Each detection terminal also has a linkage supplementary lighting device. The illumination area of the linkage supplementary lighting device matches the area where the circuit breaker is located. When the linkage supplementary lighting device emits light, the circuit breaker's opening and closing components and status indicators are illuminated. The linkage supplementary lighting device is communicatively connected to the edge computing box.
5. The remote intelligent circuit breaker opening and closing detection system according to claim 4, characterized in that: The linkage supplementary lighting device includes a photoresistor, a supplementary light, and an ADC module. The photoresistor and the supplementary light are both electrically connected to the ADC module, and the ADC module is communicatively connected to the edge computing box. The photoresistor is used to detect the ambient light intensity in the area where the circuit breaker is located, and the supplementary light is used to emit light when a supplementary light signal is received.
6. The remote intelligent circuit breaker opening and closing detection system according to claim 5, characterized in that: The ADC module is connected to the serial communication unit of the edge computing box. The ADC module has two computing modes: interrupt mode and polling mode. In interrupt mode, the edge computing box switches to ambient light processing mode by triggering a hardware interrupt, and at the same time sends a supplementary light signal to the supplementary light. In polling mode, the edge computing box receives polling requests from the edge computing box.
7. The remote intelligent circuit breaker opening and closing detection system according to claim 1, characterized in that: The remote service subsystem includes a multi-layer switch, an NVR server, a detection server, and a remote monitoring platform. The multi-layer switch connects to the edge computing box, the NVR server, and the detection server, respectively, and the detection server communicates with the remote monitoring platform.
8. The remote intelligent circuit breaker opening and closing detection system according to claim 7, characterized in that: Multilayer switches are used to provide PoE power support for image sensors and edge computing boxes in the detection terminal, while providing Ethernet network switching and VLAN isolation support for edge computing boxes, NVR servers, and detection servers.
9. The remote intelligent circuit breaker opening and closing detection system according to claim 7, characterized in that: The communication connection protocols between the multilayer switch and the edge computing box include the RTSP protocol and the WebSocket protocol. The RTSP protocol is used to transmit video streams from the image sensor transmitted through the edge computing box, and the WebSocket protocol is used to transmit circuit breaker opening and closing information from the edge computing box. The communication connection protocol between the multilayer switch and the NVR server is the RTSP protocol. The NVR server is used to provide video stream segment storage and video stream relay support between various components.
10. The remote intelligent circuit breaker opening and closing detection system according to claim 7, characterized in that: The detection server is used to feed back the circuit breaker opening and closing information obtained by the multi-layer switch to the remote monitoring platform in real time; the communication connection between the detection server and the remote monitoring platform adopts L2TP encrypted transmission connection, and the communication connection protocol used is WebSocket protocol.