Method for monitoring opening and closing speed of contact switch of gas-insulated metal-enclosed switchgear

Abstract

The invention provides a method for monitoring the opening and closing speed of a contact switch of a gas insulated metal-enclosed switchgear. The method comprises the following steps: receiving a starting signal from a disconnecting link by a routing server; the routing server sends the video image acquisition request to a video server; the video server starts to collect a disconnecting link video image; the AI algorithm server performs disconnecting link speed condition analysis on the video image acquired by the video server; the disconnecting link speed analysis result is transmitted to the application server in real time through the routing server; and the application server receives the disconnecting link speed analysis result for subsequent processing. According to the invention, when the speed is not in the preset speed range in the opening and closing process of the contact switch, the device can be judged to be abnormal so as to give an early warning to prevent safety accidents.

Description

Technical Field

[0001] The invention relates to the field of GIS equipment isolation switch monitoring, and in particular to a method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear. Background Art

[0002] Gas insulated metal enclosed switchgear and controlgear (GIS) is a metal enclosed switchgear and controlgear that uses gas at a pressure higher than atmospheric pressure as the insulating medium in at least part of the equipment. It is filled with SF6 insulating gas at a certain pressure, and is also called SF6 fully enclosed combination electrical equipment. The measurement of the opening and closing speed of the contact switch in the cavity is an important detection process. During the opening and closing process of the contact switch, if the speed is not within the predetermined speed range, it can be judged that the equipment is abnormal and an early warning can be issued to prevent safety accidents. However, there has been no good solution to the problem of measuring and monitoring the opening and closing speed of the contact. Summary of the invention

[0003] The present invention provides a contact opening and closing speed monitoring method for gas insulated metal enclosed switchgear (GIS), which aims to solve the real-time speed monitoring of the opening and closing actions of the existing GIS disconnector, so that when the speed is not within a predetermined speed range during the opening and closing process of the contact switch, it is convenient to judge that the equipment is abnormal and issue an early warning to prevent safety accidents.

[0004] In order to achieve the above object, an embodiment of the present invention provides a method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear, comprising: The routing server receives a start signal from a knife switch; the routing server sends a request to a video server; the video server starts collecting knife switch video images; the AI ​​algorithm server analyzes knife switch speed conditions based on the video images collected by the video server; the knife switch speed analysis results are transmitted to the application server in real time through the routing server; the application server receives the knife switch speed analysis results for subsequent processing.

[0005] The above structure is adopted to automatically receive the knife switch opening and closing signal, and the image is recorded and saved through the video server. The AI ​​algorithm server analyzes and processes the image to obtain the real-time detection of the knife switch opening and closing speed, which is convenient for the staff to obtain the real-time speed of the GIS tank contact opening and closing action, and to carry out preventive maintenance of abnormal equipment operation according to the real-time speed and historical speed comparison to avoid safety accidents. The AI ​​algorithm server includes an AI algorithm training module to realize the standard open state recognition of the knife switch state, the standard closed state recognition of the knife switch state, and the real-time position recognition of the knife switch.

[0006] Further: Standard opening and closing state recognition is performed through traditional algorithms.

[0007] With the above structure, the standard opening and closing actions of the contacts are collected through traditional algorithms, and after image processing, features are proposed and a model is established for opening and closing state recognition.

[0008] Furthermore: the recognition during the opening and closing process is performed by combining static recognition method and dynamic recognition method for position recognition based on the neural network target recognition algorithm YOLOV5S.

[0009] The above structure provides a recognition accuracy rate of up to 99%, thereby improving the accuracy of the system.

[0010] Further: Each service component is deployed separately.

[0011] The above structure is adopted to make the reliability of the whole detection process better.

[0012] Further: the entire component connection adopts optical fiber communication.

[0013] By adopting the above structure, the video delay is reduced to less than 1 second.

[0014] Further: Video compression adopts H.264 / H.265 real-time encoding compression technology and supports CBR and VBR.

[0015] With the above structure, the transmission bandwidth requirement can be reduced to below 500kbps.

[0016] Further: the server where the server is located is Intel E5-2620v4.

[0017] With the above structure, the multi-core performance is strong, the power consumption and heat generation are low, and the performance is good.

[0018] The above scheme of the present invention has the following beneficial effects: The present application ensures effective real-time speed monitoring of GIS equipment during the opening and closing process of the contact switch, so that if the speed is not within a predetermined speed range during the opening and closing process of the contact switch, it can be judged that the equipment is abnormal and an early warning can be issued to prevent safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a structural processing flow chart of the present invention, Figure 2 This is the principle diagram of the traditional algorithm recognition of the present invention. Figure 3 It is a recognition performance diagram of the neural network algorithm YOLOV5S of the present invention for the contact switch. DETAILED DESCRIPTION

[0020] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0021] like Figure 1 As shown, the embodiment of the present application provides a method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear, comprising the following specific steps: S111. The routing server receives a start signal from the switch; S112. The routing server sends the video image acquisition request to the video server; S113. The video server starts collecting knife switch video images; S114. The AI ​​algorithm server analyzes the speed of the knife switch based on the video images collected by the video server; S115. The switch speed analysis results are transmitted to the application server in real time through the routing server; S116. The application server receives the switch speed analysis result for subsequent processing.

[0022] The GIS contact switch opening and closing speed monitoring system based on image recognition solves the problem of single contact opening and closing state detection. By analyzing the image characteristics of the contact switch taken by the terminal camera, the dynamic and static analysis method is adopted to detect and judge the contact opening and closing distance, so as to obtain the accurate actual contact opening and closing distance, and the time is judged in combination with the number of frames of the image to obtain the time running speed. Image recognition of GIS contact switch opening and closing speed can effectively monitor the speed of switch opening and closing, ensuring the safety of life and property.

[0023] 1. Switch open and closed state recognition, switch recognition A. Standard closing and opening identification ① First, before image recognition, it is necessary to acquire the standard closing image of the closed switch and the standard opening image of the opened switch, and then grayscale process the image to convert it into a grayscale image; ② Secondly, the grayscale image of the closing state and the grayscale image of the opening state are used for difference processing to obtain a grayscale image of only the moving contact; ③Finally, after obtaining the grayscale image of the contact, the grayscale image is binarized and filtered to obtain the binary image of the contact, and then the contour of the binary image of the contact is recognized to obtain the contour of the moving contact. The contour of the current contact image is compared with the contour of the standard contact, the current contact state is obtained, and the analysis results are output. The schematic diagram is as follows Figure 2 shown.

[0024] B. Contact switch recognition: For switch recognition, we use the target detection algorithm YOLOV5S based on a deep neural network. The Yolov5 model flexibly configures models of different complexities by applying channel and layer control factors similar to EfficientNet, and adopts a cross-neighborhood grid matching strategy in the positive and negative sample definition stage to obtain more positive sample anchors and accelerate convergence. In order to test the performance of the switch detection algorithm, we collected 2,000 labeled switch data and divided them into training sets and test sets in a ratio of 7:3. The mAP value statistics are shown below: Figure 3 As shown (it can be considered that the higher the mAP, the better).

[0025] 2. Analysis of the actual opening and closing position of GIS contact switches In order to effectively identify the contact switch position, the present application provides two identification methods from static and dynamic perspectives, which will be described in detail below.

[0026] A. Static analysis method: The static analysis method is to analyze only a single frame of image, mainly using a calculation method based on a circular rectangular shield. From a static perspective, the circular rectangular shield part of the contact switch represents the opening and closing position information, so from the perspective of image vision, we only need to identify the circular rectangular shield area. After finding the circular rectangular shield part, based on the top circular rectangular shield rectangular block in the image position, the current position of the contact switch can be analyzed. However, combined with the ratio of the image occupied by this position, the actual running distance of the circular rectangular shield part is obtained.

[0027] B. Dynamic analysis method: Dynamic analysis method is mainly applicable to the scene where the contact switch changes in real time. Based on the change of the contact switch, the changed area can be identified by subtracting pixels from two consecutive time series images. This area is the position of the contact switch. If the contact switch does not change, the two consecutive time series images are consistent, so it is impossible to obtain the changed area by pixel subtraction.

[0028] 3. Calculation of opening and closing speed of GIS contact switch Speed ​​measurement formula , where △S is the ratio of the actual position of the contact switch to the overall contact switch, S1 is the actual position of the contact switch, S2 is the length of the entire contact switch shield, and the monitoring distance △S=|S2-S1| is obtained; where △T is the time interval between the two image frames from the contact being fully opened or closed to the actual monitoring position; △t is the image interval time corresponding to the camera shooting rate. The camera shooting rate is set to 40pbs, ​​that is, the camera shoots 40 frames per second, then the image interval time between two adjacent frames is 25ns. The interval time between the two images is T=25(mn)ns. In the selection of △t, a very small time interval is used to obtain the instantaneous operating speed of the contact switch. Since the time interval of the selected image sequence is small, the instantaneous speed V of the contact switch can be understood as the real-time speed of the contact.

[0029] This specific embodiment is merely an explanation of the present invention and is not a limitation of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the present embodiment as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by the patent law.

Claims

1. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear, characterized in that: include: The routing server receives a start signal from the knife switch; The routing server sends the video image acquisition request to the video server; The video server starts to collect the knife switch video images; The AI ​​algorithm server analyzes the speed of the knife switch based on the video images collected by the video server; The results of knife switch speed analysis are transmitted to the application server in real time through the routing server; The application server receives the switch speed analysis results for subsequent processing.

2. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear according to claim 1, characterized in that: The AI ​​algorithm server includes an AI algorithm training module for realizing standard open state recognition of the knife switch, standard closed state recognition of the knife switch, and real-time position recognition of the knife switch.

3. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear according to claim 2, characterized in that: The real-time position recognition of the knife switch includes static position recognition and dynamic position recognition.

4. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear according to claim 3, characterized in that: The static position recognition is performed through the YOLOV5S target recognition algorithm combined with image analysis, the distance is confirmed by the position, the time to reach the current position is calculated by confirming the number of picture frames from the standard opening and closing state to the current time point, and the real-time speed is obtained through V=S / T.

5. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear according to claim 4, characterized in that: Dynamic position recognition can identify the changed area by subtracting pixels from two consecutive time-series images, and this area is the position of the contact switch.

6. A method for monitoring the opening and closing speed of contacts of a gas-insulated metal-enclosed switchgear according to claim 1, characterized in that: The application server receives the knife switch speed analysis result for subsequent processing. Specifically, the application server receives the real-time contact opening and closing detection speed information for front-end application information display.

Images