CN-122023381-A - State monitoring method and device for array type pressing plate, terminal equipment and storage medium

Abstract

The invention discloses a state monitoring method and device of an array type pressing plate, terminal equipment and a storage medium, and belongs to the technical field of power equipment detection, wherein the method comprises the following steps: inputting the operation panel image into a preset pressure plate array recognition model, so that the pressure plate array recognition model generates a pressure plate detection frame of each pressure plate on the operation panel image, and determining the pressure plate state of each pressure plate; generating an array detection frame containing all the pressing plates according to a plurality of pressing plate detection frames, generating a plurality of grids and second row-column coordinates of each grid according to the array layout, labeling corresponding pressing plate labels for each grid according to the first row-column coordinates and the second row-column coordinates, matching the grids with the pressing plate detection frames, and associating the pressing plate labels of the successfully matched grids with the pressing plate detection frames to generate a state detection report of the array pressing plate. Therefore, the invention overcomes the defect of low monitoring efficiency in the prior art.

Inventors

• SUN WENXING
• NING XUEFENG
• LIANG YONGCHAO
• ZHANG MIN
• XU GANG
• MAI WEIHUA
• Zhang Peirun
• ZHANG JIANGUO
• YANG CHAO
• LI DUANJIAO
• QIN GUANGSHENG
• LIU JIANMING
• JIA ZIRAN
• Yao Juanwen
• WU XIAOLIANG
• YUE XIAO
• CHEN BIN

Assignees

• 广东电网有限责任公司

Dates

Publication Date

20260512

Application Date

20260225

Claims (10)

1. 1. A method for monitoring the state of an array platen, comprising: acquiring an operation panel image of an array type pressing plate and an array layout of the array type pressing plate, wherein the array layout records first row coordinates and pressing plate labels of all the pressing plates; Inputting the operation panel image into a preset pressure plate array recognition model, so that the pressure plate array recognition model recognizes each pressure plate in the operation panel image, generates a pressure plate detection frame of each pressure plate on the operation panel image, and determines the pressure plate state of the pressure plate in each pressure plate detection frame; generating an array detection frame comprising all the pressing plates according to the plurality of pressing plate detection frames; dividing the array detection frame according to the array layout, generating a plurality of grids and second row-column coordinates of each grid, and labeling each grid with a corresponding pressing plate label according to the first row-column coordinates and the second row-column coordinates; Calculating the intersection ratio of the grid and the pressure plate detection frame, matching the grid with the pressure plate detection frame according to the intersection ratio, and associating the pressure plate label of the successfully matched grid with the pressure plate detection frame; and generating a state detection report of the array type pressing plate according to the pressing plate label and the pressing plate state of each pressing plate detection frame.
2. 2. The method for monitoring the state of an array platen according to claim 1, wherein the platen array recognition model is constructed, comprising: acquiring initial images of operation panels of a plurality of array type pressing plates; labeling each pressing plate in each initial image, generating a labeling detection frame and a labeling state of each pressing plate, and constructing a training data set by taking the labeled initial image as a training image; Constructing an initial image recognition model; And training the initial image recognition model by adopting the training data set, calculating a loss function value according to a preset loss function and an output image of the initial image recognition model after each round of training is completed, adjusting model parameters of the initial image recognition model when the loss function value is determined not to be converged, and taking the initial image recognition model as the platen array recognition model when the loss function value is determined to be converged.
3. 3. The method of claim 2, wherein generating an array inspection frame including all the platens according to the plurality of platen inspection frames, comprises: Generating initial pixel coordinates of boundary points of each pressure plate detection frame according to the pressure plate detection frames; Taking the consistent size of each pressing plate detection frame as a target, performing perspective transformation on the operation panel image according to the initial pixel coordinates, and generating target pixel coordinates of each pressing plate detection frame after perspective transformation; determining a minimum abscissa, a minimum ordinate, a maximum abscissa and a maximum ordinate from a plurality of target pixel coordinates; and generating an array detection frame according to the minimum abscissa, the minimum ordinate, the maximum abscissa and the maximum ordinate.
4. 4. The method for monitoring the status of an array platen according to claim 3, wherein the dividing the array detection frame according to the array layout to generate a plurality of grids and second row-column coordinates of each grid, and labeling each grid with a corresponding platen label according to the first row-column coordinates and the second row-column coordinates comprises: determining the number of rows and the number of columns of the array pressing plate according to the row-column coordinates in the array layout; dividing the array detection frame equally according to the number of rows and the number of columns to generate a plurality of grids with consistent sizes; and matching the second row-column coordinates of each grid with the first row-column coordinates of each pressing plate in the array layout, and determining the pressing plate label of each grid.
5. 5. The method of claim 4, wherein calculating the cross-over ratio of the grid to the platen detection frame, matching the grid to the platen detection frame according to the cross-over ratio, and associating the platen label of the successfully matched grid with the platen detection frame comprises: Obtaining grid pixel coordinates of boundary points of each grid; calculating the intersection ratio of each grid and each pressure plate detection frame according to the grid pixel coordinates and the target pixel coordinates; judging whether the maximum intersection ratio is larger than a preset threshold value according to each pressing plate detection frame, if so, taking the grid corresponding to the maximum intersection ratio as a grid successfully matched, associating the pressing plate label of the grid successfully matched with the pressing plate detection frame, and if not, marking the pressing plate detection frame as unmatched.
6. 6. The method of claim 5, wherein generating the status detection report of the array platen according to the platen label and the platen status of each platen detection frame comprises: Acquiring a preset state table of the array type pressing plates, wherein the preset state table records pressing plate labels and preset states corresponding to all pressing plates in the array type pressing plates; generating an actual state table according to the pressure plate label and the pressure plate state of each pressure plate detection frame; comparing the preset state and the pressing plate state under the same pressing plate label according to the actual state table and the preset state table; marking the detection result of the corresponding pressing plate label as a state coincidence when the preset state and the pressing plate state are the same, and marking the detection result of the corresponding pressing plate label as a state anomaly when the preset state and the pressing plate state are different; and generating a state detection report according to the pressing plate label, the pressing plate state, the preset state and the detection result.
7. 7. A condition monitoring device for an array platen, comprising: The system comprises a data acquisition module, a data processing module and a data processing module, wherein the data acquisition module is used for acquiring an operation panel image of an array type pressing plate and an array layout of the array type pressing plate, wherein the array layout records first row and column coordinates of each pressing plate and a pressing plate label; the model identification module is used for inputting the operation panel image into a preset pressure plate array identification model, so that the pressure plate array identification model identifies each pressure plate in the operation panel image, generates a pressure plate detection frame of each pressure plate on the operation panel image, and determines the pressure plate state of the pressure plate in each pressure plate detection frame; the detection frame generation module is used for generating an array detection frame comprising all the pressing plates according to the plurality of pressing plate detection frames; The detection frame processing module is used for dividing the array detection frame according to the array layout, generating a plurality of grids and second row-column coordinates of each grid, and labeling corresponding pressure plate labels for each grid according to the first row-column coordinates and the second row-column coordinates; The detection frame matching module is used for calculating the intersection ratio of the grid and the pressure plate detection frame, matching the grid with the pressure plate detection frame according to the intersection ratio, and associating the pressure plate label of the successfully matched grid with the pressure plate detection frame; And the pressure plate state detection module is used for generating a state detection report of the array pressure plate according to the pressure plate label and the pressure plate state of each pressure plate detection frame.
8. 8. The apparatus for monitoring the status of an array platen according to claim 7, further comprising a platen array recognition model building module; the pressing plate array recognition model construction module is used for acquiring initial images of operation panels of a plurality of array pressing plates; labeling each pressing plate in each initial image, generating a labeling detection frame and a labeling state of each pressing plate, and constructing a training data set by taking the labeled initial image as a training image; Constructing an initial image recognition model; And training the initial image recognition model by adopting the training data set, calculating a loss function value according to a preset loss function and an output image of the initial image recognition model after each round of training is completed, adjusting model parameters of the initial image recognition model when the loss function value is determined not to be converged, and taking the initial image recognition model as the platen array recognition model when the loss function value is determined to be converged.
9. 9. A terminal device, comprising: one or more processors; A memory coupled to the processor for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement a method of condition monitoring an arrayed platen as claimed in any one of claims 1-6.
10. 10. A storage medium having a computer program stored thereon, which when executed by a processor implements a method of condition monitoring an arrayed platen according to any one of claims 1-6.

Description

State monitoring method and device for array type pressing plate, terminal equipment and storage medium Technical Field The present invention relates to the field of power equipment detection technologies, and in particular, to a method and apparatus for monitoring a state of an array platen, a terminal device, and a storage medium. Background In modern industrial control systems, in particular in power system relay protection devices, the provision of an array-type protective platen for an operating panel has become a common and important design. The design generally arranges a plurality of protection pressing plates with different functions on one panel in a concentrated way, so that a compact and visual operation interface is formed, and operation personnel can conveniently perform operations such as function switching, isolation, resetting and the like. Under the scene, the state of each protection pressing plate or button is monitored accurately in real time, and the monitoring is a key link for ensuring the safe and reliable operation of the system. However, current image recognition techniques face significant limitations when applied to this scenario. Although the technology can more reliably identify the physical state of each pressing plate through a high-resolution camera and an advanced image processing algorithm, for example, whether the pressing plate is in a 'put-in' position or an 'withdraw' position, the 'split' state or the 'combine' state or the color identification can be effectively extracted. However, when faced with an array of closely spaced, similarly looking, numerous platens, the bottleneck of the prior art is evident by the lack of the ability to accurately match the identified status information to each specific platen identity in the array. In short, the system can judge that a plurality of pressing plates are in a 'put-in' state in a picture, but the system cannot automatically and reliably judge whether the pressing plates respectively correspond to an overcurrent protection I-section pressing plate, a differential protection pressing plate, a reclosing pressing plate and the like. This lack of matching capability results in operators still having to rely on manual checks to confirm the status of a particular platen, and thus, current array platen status monitoring methods suffer from the disadvantage of low monitoring efficiency. Disclosure of Invention The invention provides a state monitoring method, device, terminal equipment and storage medium for an array type pressing plate, and the method can solve the defect of low monitoring efficiency in the prior art. An embodiment of the present invention provides a method for monitoring a state of an array platen, including: acquiring an operation panel image of an array type pressing plate and an array layout of the array type pressing plate, wherein the array layout records first row coordinates and pressing plate labels of all the pressing plates; Inputting the operation panel image into a preset pressure plate array recognition model, so that the pressure plate array recognition model recognizes each pressure plate in the operation panel image, generates a pressure plate detection frame of each pressure plate on the operation panel image, and determines the pressure plate state of the pressure plate in each pressure plate detection frame; generating an array detection frame comprising all the pressing plates according to the plurality of pressing plate detection frames; dividing the array detection frame according to the array layout, generating a plurality of grids and second row-column coordinates of each grid, and labeling each grid with a corresponding pressing plate label according to the first row-column coordinates and the second row-column coordinates; Calculating the intersection ratio of the grid and the pressure plate detection frame, matching the grid with the pressure plate detection frame according to the intersection ratio, and associating the pressure plate label of the successfully matched grid with the pressure plate detection frame; and generating a state detection report of the array type pressing plate according to the pressing plate label and the pressing plate state of each pressing plate detection frame. Further, the construction of the platen array recognition model includes: acquiring initial images of operation panels of a plurality of array type pressing plates; labeling each pressing plate in each initial image, generating a labeling detection frame and a labeling state of each pressing plate, and constructing a training data set by taking the labeled initial image as a training image; Constructing an initial image recognition model; And training the initial image recognition model by adopting the training data set, calculating a loss function value according to a preset loss function and an output image of the initial image recognition model after each round of training is completed, adjusting m