Power equipment nameplate information identification and verification system and method based on regulation and control cloud

Abstract

The invention discloses a power equipment nameplate information identification and verification system and method based on a regulation and control cloud, and the method comprises the steps: collecting a nameplate image of secondary equipment, and uploading the nameplate image to a corresponding database of the regulation and control cloud; designing a nameplate template of the related nameplate image; preprocessing the acquired nameplate image to obtain a preprocessed image; the preprocessed image is input to an OCR processing interface, an OCR engine is selected to complete recognition of text content, and the recognized text content is classified according to a nameplate template; comparing the pictures identified and classified by the OCR engine with the secondary equipment ledger information in the regulation and control cloud system, and searching and marking similarities and differences between the pictures and the secondary equipment ledger information; after machine account comparison is completed, corresponding data in the regulation and control cloud system are modified according to machine account errors or omission found by the system, and operation and maintenance personnel are recorded and prompted to modify content. According to the invention, the problems of low efficiency, high error rate and the like of traditional manual input are effectively solved, and intelligent upgrading of power equipment management is realized.

Description

Technical Field

[0001] The present invention relates to the technical fields of nameplate recognition detection, text analysis, etc. of a power grid regulation cloud system, and specifically relates to a power equipment nameplate information recognition and verification system and method based on the regulation cloud. Background Art

[0002] There are problems in the State Grid system such as inaccurate information, untimely updates, and difficult maintenance in the secondary equipment management system of the regulation cloud. The equipment nameplate records the basic parameters of the power grid secondary equipment, including equipment model, manufacturer, number of network ports, number of power supplies, etc. Through the intelligent analysis of the on-site photos of the equipment, the project realizes the extraction of various key parameters of the equipment from the nameplate, panel, and power supply photos, and automatically compares and verifies them with the equipment account data in the regulation cloud system, forming an accurate and real-time equipment account maintenance process.

[0003] The existing equipment nameplate recognition only stays at recognizing the nameplate information and has not been studied deeply. The existing mainstream nameplate recognition methods are divided into two categories. One category is to use the OCR (Optical Character Recognition) method, using OCR software or development kits to realize the character extraction of the nameplate pictures. The other category is to use image processing methods to realize the character extraction using image processing algorithms. However, the existing methods fail to further analyze the equipment account information and cannot help the operation and maintenance personnel to determine whether there are omissions or errors in the equipment account information. Summary of the Invention

[0004] The present invention provides a power equipment nameplate information recognition and verification system and method based on the regulation cloud, which automatically extracts the account information in the on-site equipment photos through OCR technology, and uses NLP technology to intelligently compare it with the existing data in the database, identify data errors or omissions, and finally realize the automatic verification and update of the account information, improving the accuracy and efficiency of power equipment management. This method effectively solves the problems of low efficiency and high error rate existing in traditional manual entry, and realizes the intelligent upgrade of power equipment management.

[0005] A power equipment nameplate information recognition and verification method based on the regulation cloud includes:

[0006] Step S1, select the secondary equipment in the secondary equipment maintenance management system of the regulation cloud, take the corresponding nameplate image according to the substation where the secondary equipment is located and the equipment name, and upload the taken nameplate image to the corresponding database of the regulation cloud;

[0007] Step S2, design a nameplate template for the relevant nameplate image according to the layout and text content of the secondary equipment nameplate;

[0008] Step S3: Preprocess the collected nameplate image to obtain the preprocessed image;

[0009] Step S4: Input the preprocessed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, complete the recognition of the text content, and classify the recognized text content according to the nameplate template;

[0010] Step S5: Compare the image recognized and classified by the OCR engine with the secondary equipment ledger information in the regulation and control cloud system, and find and mark the similarities and differences between the two;

[0011] Step S6: After completing the ledger comparison, modify the corresponding data in the regulation and control cloud system according to the ledger errors or omissions found by the system, and record and prompt the maintenance personnel of the modification content.

[0012] Further, the secondary equipment includes a telecontrol device, a phasor measurement unit, an electric energy acquisition terminal, a router, a switch, a longitudinal encryption device, a transverse isolation device, or a server.

[0013] Further, the nameplate image includes a photo of the front panel of the device, a photo of the rear panel of the device, a photo of the device nameplate, or a photo of the device power supply.

[0014] Further, step S2 specifically includes:

[0015] Step S21: First, divide the content area of the nameplate image, disassemble the content according to the text content and layout format on the nameplate, and then define the corresponding category string according to the text content on the nameplate;

[0016] Step S22: Design a content text box for storing the device fields in the secondary equipment maintenance management system of the regulation and control cloud, and number the different strings according to the order of the text recognized by the text recognition system.

[0017] Further, step S3 specifically includes:

[0018] S31: Perform grayscale processing on the image. Set a grayscale processing function in the code, input the original image, convert the color image collected on-site into a grayscale image, and output the converted grayscale image;

[0019] S32: Perform binary processing on the image. Set a binary function in the code, input the grayscale image, further convert the grayscale image into a binary image, and output the binary image;

[0020] S33: Perform denoising processing on the image. Input the binary image, and use methods such as median filtering and Gaussian filtering to remove the interference information on the binary image, and output the binary image after the filtering operation;

[0021] S34, The input is a binary image that has undergone a filtering operation. Use the Hough transform and affine transform to rotate the image and adjust the image orientation to ensure that the text is horizontally arranged. The output is a binary image after rotation;

[0022] S35, Perform edge detection. The input is the binary image after rotation. Use the Canny edge detection algorithm to extract the contour of the document. The output is a binary image marking the edges of the text;

[0023] S36, Perform Gaussian blur processing. The input is a binary image. Use Gaussian blur operation to reduce the high-frequency noise in the image. The output is an image with reduced image noise;

[0024] S37, Resize the image. The input is the image with reduced image noise. Adjust the aspect ratio of the image. The output is an image sized appropriately for OCR engine processing.

[0025] Further, step S4 specifically includes:

[0026] S41, Environment configuration. Configure the environment required for OCR algorithm implementation: Select the OCR engine, Tesseract, and configure the parameters of the OCR engine, including setting the languages to Chinese and English, the recognition mode to the default mode, and the confidence interval to 90;

[0027] S42, Text recognition. Input the nameplate image to be recognized and call the OCR engine to complete image recognition and text content extraction;

[0028] S43, Text matching. Based on the nameplate template and the text content obtained by OCR engine recognition, divide the text. Divide different texts into different categories according to their corresponding positions; According to the template layout and image content, divide the text content on the image into different types of text boxes; Based on the comparison and judgment of the text positions on the template and the collected image, divide the text content recognized on the nameplate into several different types of text boxes.

[0029] Further, step S5 specifically includes:

[0030] S51, Text classification. Classify the text according to the recognized text content in the order output by the system and the marked text categories, and associate it with the corresponding server ledger content in the secondary equipment database of the regulation cloud system;

[0031] S52. Data default prompt: When the device parameters obtained by recognition do not exist in the account information of the corresponding device in the secondary device database of the regulation cloud system, mark and prompt this item of data; when this item of device parameter exists in the secondary device database of the regulation cloud system but is not recognized by the system, mark and prompt this item of parameter; when there is the same field in both the database and the nameplate recognition content, but the corresponding content of the field is empty, mark and prompt this item of parameter.

[0032] S53. Numerical comparison: When the parameter content on the recognized device nameplate is a number, compare it with the corresponding content in the secondary device database of the regulation cloud system. When the two numerical values are inconsistent, mark and prompt this item of data.

[0033] S54. Text comparison: When the parameter content on the recognized device nameplate is text, use NLP technology to analyze the text content. When the text content is inconsistent with the secondary device database of the regulation cloud system, mark and prompt this item of data.

[0034] Further, step S6 specifically includes:

[0035] S61. Automatic filling of default data: When the corresponding data exists in the secondary device database of the regulation cloud system but is not recognized by the system, prompt the relevant personnel; when the corresponding data does not exist in the secondary device database of the regulation cloud system but is recognized by the system, the system automatically fills it in the database according to the recognized content.

[0036] S62. Numerical modification: When the device parameters obtained by recognition are inconsistent with the device parameters in the secondary device database of the regulation cloud system and this parameter is of numerical type, the system automatically modifies the parameter information according to the recognized content and automatically records the modification content.

[0037] S63. Text modification: When the parameter content on the recognized device nameplate is text and is inconsistent with the corresponding content in the secondary device database of the regulation cloud system, replace the text in the database according to the recognized text content and automatically record the modification content.

[0038] Further, it also includes:

[0039] Step S7. The regulation personnel set a trigger. When the trigger meets the set time and event requirements, the system automatically conducts the verification of the secondary device account information in the regulation cloud system, automatically completes the relevant functions of steps S5 and S6, and stores the verification results and updated content in the system log. The trigger includes a time trigger and a time trigger.

[0040] The regulation personnel set time triggers according to production requirements to achieve automatic inspection of the system. In the time triggers, the regulation personnel set the daily automatic inspection times and the time for each inspection. When the system time reaches the set time, the system automatically starts the inspection;

[0041] The regulation personnel set event triggers according to production requirements to achieve automatic inspection of the system. The trigger events set by the event triggers include: equipment inspection timeout, the number of inspection error reports exceeding the threshold, and the completion of equipment ledger information update. When the set trigger events occur, the event trigger is automatically triggered to conduct another automatic inspection of the system.

[0042] A power equipment nameplate information recognition and verification system based on the regulation cloud is applied to the method described above. The system includes:

[0043] An image acquisition module, which is used to select secondary equipment in the secondary equipment maintenance and management system of the regulation cloud, take corresponding nameplate images according to the plant station and equipment name of the secondary equipment, and upload the taken nameplate images to the corresponding database of the regulation cloud;

[0044] A template definition module, which is used to design a nameplate template for the relevant nameplate images according to the layout and text content of the secondary equipment nameplate;

[0045] An image preprocessing module, which is used to preprocess the acquired nameplate images to obtain preprocessed images;

[0046] An OCR recognition module, which is used to input the preprocessed image into the OCR processing interface, select an OCR engine, configure relevant parameters and variables, complete the recognition of the text content, and classify the recognized text content according to the nameplate template;

[0047] A ledger comparison module, which is used to compare the images recognized and classified by the OCR engine with the secondary equipment ledger information in the regulation cloud system, and find and mark the similarities and differences between the two;

[0048] A ledger modification module, which is used to modify the corresponding data in the regulation cloud system according to the ledger errors or omissions found by the system after completing the ledger comparison, and record and prompt the maintenance personnel of the modification content.

[0049] The present invention has the following beneficial effects and characteristics:

[0050] 1. The power equipment nameplate information recognition and verification method and system based on the regulation cloud of the present invention mainly uses OCR technology to realize the recognition and extraction of the nameplate information of secondary equipment managed by the regulation cloud, and combines natural language processing methods to realize the verification and automatic update of the content extracted from the equipment nameplate and the data in the database. The present invention is used to realize the rapid collection, management, verification, etc. of the secondary equipment ledger information in the regulation cloud system, and assist the operation and maintenance personnel in data entry and fault handling;

[0051] 2. The present invention can automatically associate and analyze the pictures of the nameplates of secondary equipment in the regulation cloud system collected on site and the information in the regulation cloud secondary equipment ledger database associated therewith, and realize the automatic update and maintenance of the secondary equipment ledger through the collection and analysis functions provided by the system.

[0052] 3. The present invention has the function of automatically processing the nameplate images of the collected secondary equipment. The present invention uses relevant algorithms for image preprocessing to improve the image quality and further improve the accuracy of text extraction. The present invention uses operations such as image grayscale conversion, image binarization, image denoising, image skew correction, edge detection, and image resizing to achieve effects such as removing background noise information, enhancing the contrast between text and background, detecting text edges, and resizing images. Through operations such as combination and arrangement of the above functions, custom processing of the collected images is realized to obtain high-quality secondary nameplate images, providing high-quality image data for the OCR text recognition function of the system.

[0053] 4. The present invention can also realize functions such as automatic verification, prompting, and log management of the regulation cloud secondary equipment ledger management system. After the ledger update function is completed, the present invention can automatically display the ledger update content in the window and automatically store the update content, update time, etc. in the system log, facilitating the operation and maintenance personnel to view the content of the secondary equipment ledger update in real time. The present invention can set an automatic inspection function. The system of the present invention can set a trigger. When the trigger meets the set time and event requirements, the system automatically conducts the verification of the secondary equipment ledger information in the regulation cloud system and stores the verification results and update content in the system log. Description of the Drawings

[0054] Figure 1 is the flowchart of a method for recognizing and verifying power equipment nameplate information based on the regulation cloud according to an embodiment of the present invention;

[0055] Figure 2 is the schematic diagram of the image preprocessing process according to an embodiment of the present invention;

[0056] Figure 3 is the flowchart of data anomaly marking according to an embodiment of the present invention;

[0057] Figure 4 The nameplate image collected in the embodiment of the present invention

[0058] Figure 5 The image after preprocessing in the embodiment of the present invention;

[0059] Figure 6 The result of OCR recognition in the embodiment of the present invention. Detailed implementation manners

[0060] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0061] Please refer to Figure 1 , the embodiment of the present invention provides a method for identifying and verifying the nameplate information of power equipment based on a regulation cloud, including the following steps:

[0062] Step 1. Image acquisition

[0063] For the scope of secondary equipment managed by the regulation cloud system, the embodiment of the present invention can collect and process the nameplate images of all secondary equipment managed by the regulation cloud system, including but not limited to the following secondary equipment: telecontrol device, phasor measurement unit, electric energy acquisition terminal, router, switch, longitudinal encryption device, transverse isolation device, server, etc.

[0064] Scope of secondary nameplate images: The embodiment of the present invention can collect various nameplate images of each secondary equipment, including but not limited to the following nameplate images: front panel photo of the equipment, rear panel photo of the equipment, nameplate photo of the equipment, power supply photo 1 of the equipment, power supply photo 2 of the equipment, etc.

[0065] In the embodiment of the present invention, a mobile phone, a camera and other digital terminals can be used as the shooting instrument to separately or batch obtain images to complete the shooting of the nameplate images of secondary equipment in the substation.

[0066] When uploading the image, log in to the regulation cloud system and search for the device according to the device name, device ID, affiliated substation, etc. Click the upload icon on the device detailed information interface to upload the nameplate image of the secondary equipment corresponding to the icon.

[0067] Step 2. Template design

[0068] When extracting the text information on the nameplates of secondary equipment in the present invention, it is necessary to first design different nameplate templates according to the nameplate layouts of different equipment, so as to record the ledger information required for different equipment and the positions of the corresponding texts on the nameplates. The present invention needs to count the types of secondary equipment and design separate templates for each type of secondary equipment to store and record the extracted equipment information.

[0069] 1) Nameplate parsing: Divide the content area of the nameplate image of the secondary equipment, specifically disassemble the content according to the text content and layout format on the nameplate, and then define the corresponding category strings according to the text content on the nameplate, such as: equipment name, equipment model, etc.

[0070] 2) Template definition: Design content text boxes to store string content excluding field names, and number different strings according to the order of the characters recognized by the character recognition system.

[0071] Step Three: Image preprocessing

[0072] The original image has defects such as rotation, blurring, and reflection, and the original image needs to be further processed to obtain a high-quality secondary nameplate image. The present invention preprocesses the collected nameplate image to obtain the preprocessed image, as Figure 2 shown, and the specific steps are as follows:

[0073] 1) Image grayscale processing: Set the image grayscale processing interface, input the original image, convert the color image collected on-site into a grayscale image, and output the converted grayscale image.

[0074] 2) Image binarization processing: Set the image binarization interface, input the grayscale image, further convert the grayscale image into a binary image, and output the binary image.

[0075] 3) Image denoising processing: Set the image denoising interface, input the binary image, and use methods such as median filtering and Gaussian filtering to remove the interference information on the binary image, and output the binary image after the filtering operation.

[0076] 4) Image skew correction: Set the image skew correction interface, adjust the image orientation to ensure that the text is horizontally arranged. The input is the binary image after the filtering operation, and use Hough transform, affine transform, etc. to rotate the image, and output the binary image after rotation.

[0077] 5) Edge detection: Set the text edge detection interface, input the image obtained through the above processing, and use algorithms such as Canny edge detection to extract the contour of the document, and output the image marked with the text edge.

[0078] 6) Gaussian Blur Processing: Set up the Gaussian blur processing interface. The input is a binary image. Use the Gaussian blur operation to reduce the high-frequency noise in the image. The output is an image with reduced image noise.

[0079] 7) Image Size: Set up the image processing interface. The input is the image processed as above. Adjust the aspect ratio of the image. The output is an image with a size suitable for processing by the OCR engine.

[0080] The image preprocessing function includes, but is not limited to, the above image preprocessing steps, and the preprocessing steps in the preprocessing stage can be screened, increased, or decreased according to actual needs.

[0081] Step Four: OCR Recognition

[0082] After completing the image preprocessing work, carry out the OCR recognition work. Input the preprocessed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, and complete the text extraction.

[0083] 1) Environment Configuration: Configure the environment required for the implementation of the OCR algorithm, select the OCR engine, such as Tesseract, PaddleOCR, etc. Configure the parameters of the OCR engine, such as system environment variables like language settings, recognition mode, confidence interval, etc.

[0084] 2) Template Matching: After the system obtains the uploaded nameplate image, it automatically matches the nameplate template defined by the upload operation according to the type of secondary equipment uploaded. The nameplate template obtained by this operation is the basic function for storing and analyzing the device text information obtained by subsequent OCR recognition.

[0085] 3) Text Recognition: Input the nameplate image to be recognized, call the OCR engine and install the predefined model parameters to complete the text recognition and content extraction on the secondary nameplate image.

[0086] 4) Text Matching: According to the text content obtained by the secondary nameplate template and OCR engine recognition, match the text, and divide different texts into different categories according to the order and text type obtained by recognition.

[0087] Furthermore, this function inputs the recognized text into the ledger comparison system to complete the matching and identification with the data in the secondary equipment ledger database in the dispatching cloud system.

[0088] Step Five: Ledger Comparison

[0089] This function compares the secondary equipment nameplate information recognized and classified by the OCR engine with the secondary equipment ledger information in the dispatching cloud system, and finds and marks the similarities and differences between the two for the convenience of the operation and maintenance personnel to carry out their work.

[0090] 1) Text classification: Classify the text according to the recognized text position and associate it with the corresponding equipment ledger content in the secondary equipment database of the regulation cloud system. This function classifies and marks the obtained secondary equipment ledger fields, including but not limited to the following fields: equipment model, number of power supplies, number of network ports, number of device Us, device U positions, etc.

[0091] 2) Data default prompt: When the recognized device parameters do not exist in the ledger information of the corresponding device in the secondary equipment database of the regulation cloud system, mark and prompt this item of data. When this item of device parameter exists in the secondary equipment database of the regulation cloud system but is not recognized by the system, mark and prompt this parameter.

[0092] 3) Numerical comparison: When the parameter content on the recognized device nameplate is a number, compare it with the corresponding content in the secondary equipment database of the regulation cloud system. When the two numerical values are inconsistent, mark and prompt this item of data:

[0093] 4) Text comparison: When the parameter content on the recognized device nameplate is text, use NLP technology to analyze the text content. When the text content is inconsistent with the secondary equipment database of the regulation cloud system, mark and prompt the inconsistent content of this item of data.

[0094] Step Six: Ledger update

[0095] After completing the ledger comparison, modify the corresponding data in the regulation cloud system according to the ledger errors or omissions found by the system, and record and prompt the maintenance personnel of the modification content.

[0096] 1) Automatic filling of default data: When the corresponding data exists in the secondary equipment database of the regulation cloud system but is not recognized by the present invention, prompt the relevant personnel. When the corresponding data does not exist in the secondary equipment database of the regulation cloud system but is recognized by the system, the system automatically completes the filling in the database according to the recognized content.

[0097] 2) Numerical modification: When the recognized device parameters are inconsistent with the device parameters in the secondary equipment database of the regulation cloud system and this parameter is of numerical type, the system automatically modifies the parameter information according to the recognized content and automatically records the modification content.

[0098] 3) Text modification: When the parameter content on the recognized device nameplate is text and is inconsistent with the corresponding content in the secondary equipment database of the regulation cloud system, replace the text in the database according to the recognized text content and automatically record the modification content.

[0099] 4) Log update: When the system completes the update of the equipment ledger in the regulation cloud system, the system automatically records information such as the update content and update time.

[0100] 5) Automatic update: The operation and maintenance personnel can set triggers. When the triggers meet the settings such as time and events, the system can automatically start the update of the secondary equipment ledger in the regulation cloud system and complete the recording and storage of system logs.

[0101] Correspondingly, the embodiment of the present invention further provides a power equipment nameplate information recognition and verification system based on the regulation cloud, including: an image acquisition module, a template design module, an image preprocessing module, an OCR recognition module, a ledger comparison module, and a ledger modification module.

[0102] The image acquisition module is connected to the template design module, the template design module is connected to the image preprocessing module, the image preprocessing module is connected to the OCR recognition module, the OCR recognition module is connected to the comparison and ledger comparison module, and the ledger comparison module is connected to the update module. For any secondary nameplate image collected on site, it needs to go through image acquisition, template matching, image preprocessing, OCR recognition, ledger comparison, and ledger update operations in sequence. The prompt methods after ledger update include pop-up windows, app push, etc.

[0103] Further, the system can set triggers to realize the automatic verification of all secondary equipment on the secondary equipment maintenance and management system of the regulation cloud system. The present invention can set trigger time, trigger events, etc. When the system detects that the trigger conditions are met, the system starts the automatic verification of secondary equipment. After all secondary equipment verifications are completed, the system automatically stores the update log.

[0104] The image acquisition module is used to select secondary equipment in the secondary equipment maintenance and management system of the regulation cloud, take corresponding nameplate images according to the plant substation and equipment name where the secondary equipment is located, and upload the taken nameplate images to the corresponding database in the regulation cloud;

[0105] The template definition module is used to design a nameplate template for relevant nameplate images according to the layout and text content of the secondary equipment nameplate, so that the text content recognized later can be further classified and associated with relevant information.

[0106] The image preprocessing module is used to preprocess the collected nameplate images to obtain preprocessed images.

[0107] The OCR recognition module is used to input the preprocessed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, complete the recognition of text content, and classify the recognized text content according to the nameplate template.

[0108] The ledger comparison module is used to compare the images recognized and classified by the OCR engine with the secondary equipment ledger information in the regulation cloud system, and find and mark the similarities and differences between the two.

[0109] The ledger modification module is used to complete the ledger comparison, modify the corresponding data in the regulation cloud system according to the ledger errors or omissions found by the system, record and prompt the maintenance personnel of the modification content.

[0110] The following takes the server as an example to elaborate in detail on the method for identifying and verifying the nameplate information of power equipment based on the regulation cloud of the present invention, including the following steps:

[0111] Step S1, select the secondary equipment in the regulation cloud secondary equipment maintenance and management system: server, take the corresponding computer motherboard image (nameplate image) according to the plant station and equipment name of the server, and upload the captured nameplate image to the corresponding database of the regulation cloud. For this step, it is necessary to select and take the motherboard image of the server. When uploading the image, log in to the regulation cloud system and retrieve the device according to the device name, device ID, affiliated plant station, etc., click the upload icon on the device details interface, and upload the image of the secondary equipment nameplate type corresponding to the icon. The original image collected is as Figure 4 shown.

[0112] Step S2, after completing the upload of the relevant server nameplate image, design the template of the relevant nameplate image.

[0113] Based on the content and layout of the text on the server nameplate, design the corresponding nameplate template for the equipment nameplate, and classify the content on the nameplate through the nameplate template to facilitate the subsequent work. For this process, more specifically, it includes the following sub-steps:

[0114] Step S21, first divide the content area of the nameplate image of the server, specifically disassemble the content according to the text content and layout format on the nameplate. Subsequently, define the corresponding category strings according to the text content on the nameplate, such as: equipment model, input voltage / current, etc.

[0115] Step S22, design the content text box for storing the device fields in the regulation cloud secondary equipment maintenance and management system, such as: device name, device model, CPU model, etc. words, to store and identify the string content recognized by the system, and number the different strings according to the order of the words recognized by the character recognition system.

[0116] Step S3, after completing the server image acquisition and nameplate template definition, carry out the image preprocessing work. The original image has defects such as rotation, blur, and reflection, and the original image needs to be further processed. Preprocess the captured nameplate image to obtain the preprocessed image. For this process, it specifically includes the following multiple sub-steps.

[0117] S31. Grayscale the image. Set a grayscale processing function in the code. Input the original image, convert the color image captured on-site into a grayscale image, and output the converted grayscale image.

[0118] S32. Binarize the image. Set a binarization function in the code. Input the grayscale image, further convert the grayscale image into a binary image, and output the binary image.

[0119] S33. Denoise the image. Input the binary image and use methods such as median filtering and Gaussian filtering to remove the interference information on the binary image. Output the binary image after the filtering operation.

[0120] S34. Rectify the inclination of the image, adjust the image orientation to ensure that the text is horizontally arranged. Input the binary image after the filtering operation, and use methods such as Hough transform and affine transform to rotate the image. Output the binary image after rotation.

[0121] S35. Perform edge detection. Input the image obtained through the above processing, and use algorithms such as Canny edge detection to extract the contour of the document. Output the binary image marking the edges of the text.

[0122] S36. Perform Gaussian blur processing. Input the binary image marking the edges of the text, and use Gaussian blur operation to reduce the high-frequency noise in the image. Output the image with reduced image noise.

[0123] S37. Resize the image. Input the image obtained through the above processing, adjust the aspect ratio of the image, and output the image with a size suitable for the OCR engine to process.

[0124] Step S3 includes but is not limited to the above image preprocessing steps, and the preprocessing steps in the preprocessing stage can be selected, added, or reduced according to actual needs. The image after preprocessing is as Figure 5 shown.

[0125] Step S4. After completing the image preprocessing work, carry out the OCR recognition work. Input the preprocessed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, complete the text extraction, and classify the recognized text content according to the nameplate template, as Figure 6 shown. Step S4 specifically includes the following multiple sub-steps.

[0126] S41. Environment configuration. Configure the environment required for the implementation of the OCR algorithm. Select the OCR engine, Tesseract, and configure the parameters of the OCR engine, such as setting the language to Chinese and English (chi_sim + eng), the recognition mode to the default mode (3), and the confidence interval to 90, etc. system environment variables.

[0127] S42, Character recognition. Input the nameplate image to be recognized and call the OCR engine to complete image recognition and text content extraction.

[0128] S43, Text matching. According to the nameplate template and the text content obtained by OCR engine recognition, divide the text. Divide different texts into different categories according to their corresponding positions. According to the template layout and image content, divide the text content on the image into different types of text boxes. Based on the comparison and judgment of the text positions on the template and the collected image, divide the recognized text content on the nameplate into several different types of text boxes.

[0129] Step S5. After completing OCR recognition, carry out the comparison work of the account books. Compare the pictures recognized and classified by the OCR engine with the secondary equipment account book information in the regulation cloud system, and find and mark the similarities and differences between the two for the convenience of the operation and maintenance personnel to carry out their work. Specifically, the results of the account book comparison in the present invention are divided into 6 types, namely, the nameplate text and the database field are exactly the same, the nameplate text recognition is a null value, the database field is a null value, both the nameplate text and the database field are null values, the nameplate text and the corresponding numerical content in the database do not match, and the nameplate text and the corresponding text content in the database do not match. Step S5 specifically includes the following multiple sub-steps (as Figure 3 shown):

[0130] S51, Text classification. Classify the text according to the recognized text content in the order output by the system and the marked text categories, and associate it with the corresponding server account book content in the secondary equipment database of the regulation cloud system.

[0131] S52, Data default prompt. When the recognized device parameters do not exist in the account book information of the corresponding device in the secondary equipment database of the regulation cloud system, mark and prompt this item of data. When this item of device parameter exists in the secondary equipment database of the regulation cloud system but is not recognized by the system, mark and prompt this item of parameter. When there is the same field in both the database and the nameplate recognition content, but the corresponding content of the field is empty, mark and prompt this item of parameter.

[0132] S53, Numerical comparison. When the parameter content on the recognized device nameplate is a number, compare it with the corresponding content in the secondary equipment database of the regulation cloud system. When the two numerical values are inconsistent, mark and prompt this item of data;

[0133] S54, Text comparison. When the parameter content on the recognized device nameplate is text, use NLP technology to analyze the text content. When the text content is inconsistent with the secondary equipment database of the regulation cloud system, mark and prompt this item of data.

[0134] Step S6, after completing the ledger comparison, modify the corresponding data in the regulation cloud system according to the ledger errors or omissions found by the system, record and prompt the maintenance personnel of the modification content. Step S6 specifically includes the following multiple sub-steps.

[0135] S61, default data automatic filling. When the corresponding data exists in the secondary equipment database of the regulation cloud system but the system fails to recognize it, prompt the relevant personnel. When the corresponding data does not exist in the secondary equipment database of the regulation cloud system but the system recognizes it, the system automatically completes the filling in the database according to the recognized content.

[0136] S62, numerical value modification. When the recognized device parameters are inconsistent with the device parameters in the secondary equipment database of the regulation cloud system and the parameter is of numerical type, the system automatically modifies the parameter information according to the recognized content and automatically records the modification content.

[0137] S63, text modification. When the parameter content on the recognized device nameplate is text and is inconsistent with the corresponding content in the secondary equipment database of the regulation cloud system, replace the text in the database according to the recognized text content and automatically record the modification content.

[0138] Step S7, the regulation personnel can set a trigger. When the trigger meets the set time and event requirements, the system automatically conducts the verification of the secondary equipment ledger information in the regulation cloud system, automatically completes the relevant functions of steps S5 and S6, and stores the verification results and updated content in the system log. Step S7 specifically includes:

[0139] S71, time trigger. The regulation personnel can set a time trigger according to production requirements to achieve automatic system inspection. The regulation personnel can set the number of daily automatic inspections and the time of each inspection. When the system time reaches the set time, the system automatically starts the inspection.

[0140] S72, event trigger. The regulation personnel can set an event trigger according to production requirements to achieve automatic system inspection. The regulation personnel can set event triggers, such as: equipment inspection timeout, the number of inspection error reports exceeding the threshold, the completion of equipment ledger information update, etc. When the set event occurs, the trigger is automatically triggered and the automatic inspection of the system is carried out again.

[0141] Step S8, after completing the system inspection, automatically generate a system log according to the inspection situation and prompt the maintenance personnel of the inspection situation of this time. The system log automatically generated by this system can set the storage path and storage time to facilitate the query of the system log and problem location. When data mismatch and data modification occur during the system inspection, the system will prompt the regulation personnel of the data situation and modification content in the form of pop-up windows, text messages, etc. after the inspection ends, and this part of information should be consistent with the content on the system log.

[0142] The embodiments of the present invention can bring the following technical effects:

[0143] 1. A method for identifying and verifying the nameplate information of power equipment based on a control cloud according to an embodiment of the present invention can effectively identify the text content on the secondary nameplate image of the dispatching cloud system, and can automatically complete the data comparison and update with the corresponding database in the dispatching cloud system.

[0144] 2. For different layouts and contents of secondary equipment nameplates, define nameplate templates to facilitate further classification of the text information on the equipment nameplates obtained by the system.

[0145] 3. The embodiments of the present invention preprocess the color images collected on site to eliminate the defects of the original images, facilitating the development of text recognition work.

[0146] 4. The embodiments of the present invention automatically realize the association and comparison between the secondary equipment ledger information on the dispatching cloud system and the text information on the secondary nameplate image recognized by the system.

[0147] 5. This embodiment can complete functions such as automatic inspection, comparison, update, and prompt of the equipment ledger information on the dispatching control cloud system.

[0148] As mentioned above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A method for identifying and verifying nameplate information of electric power equipment based on control cloud, characterized in that: The steps include: Step S1, select the secondary equipment in the secondary equipment maintenance management system of the control cloud, take the corresponding nameplate image according to the plant and equipment name of the secondary equipment, and upload the taken nameplate image to the corresponding database of the control cloud; Step S2, designing a nameplate template of a related nameplate image according to the layout and text content of the secondary equipment nameplate; Step S3, preprocessing the collected nameplate image to obtain a preprocessed image; Step S4, input the pre-processed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, complete the recognition of text content, and classify the recognized text content according to the nameplate template; Step S5, comparing the image identified and classified by the OCR engine with the secondary equipment ledger information in the control cloud system, finding and marking the similarities and differences between the two; Step S6, after the comparison of the ledgers is completed, the corresponding data in the control cloud system is modified according to the errors or omissions in the ledgers found by the system, and the modification content is recorded and prompted to the operation and maintenance personnel.

2. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: The secondary equipment includes a telecontrol device, a phasor measurement device, an electric energy collection terminal, a router, a switch, a longitudinal encryption device, a transverse isolation device or a server.

3. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: The nameplate image includes a photo of a front panel of a device, a photo of a rear panel of a device, a photo of a nameplate of a device, or a photo of a power supply of a device.

4. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Step S2 specifically includes: Step S21, firstly divide the content area of ​​the nameplate image, decompose the content according to the text content and typesetting format on the nameplate, and then define the corresponding category character string according to the text content on the nameplate; Step S22, design a content text box for storing equipment fields in the control cloud secondary equipment maintenance management system, and number different character strings according to the order of text obtained by the text recognition system.

5. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Step S3 specifically includes: S31, graying the image, setting a graying processing function in the code, inputting the original image to convert the color image collected on site into a graying image, and outputting the converted graying image; S32, binarizing the image, setting a binarization function in the code, inputting a grayscale image, further converting the grayscale image into a binary image, and outputting a binary image; S33, performing image denoising, inputting a binary image, removing interference information on the binary image by using methods such as median filtering and Gaussian filtering, and outputting a binary image after filtering; S34, inputting the binary image after the filtering operation, rotating the image using Hough transformation and affine transformation, adjusting the image direction, ensuring that the text is arranged horizontally, and outputting the rotated binary image; S35, performing edge detection, inputting the rotated binary image, extracting the outline of the document using the Canny edge detection algorithm, and outputting a binary image with the edges of the marked text; S36, performing Gaussian blur processing, the input is a binary image, the high-frequency noise in the image is reduced by using the Gaussian blur operation, and the output is an image with reduced image noise; S37, adjusting the image size, inputting an image with reduced image noise, adjusting the image aspect ratio, and outputting an image with a size suitable for processing by the OCR engine.

6. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Step S4 specifically includes: S41, environment configuration, configure the environment required for the OCR algorithm implementation: select the OCR engine, Tesseract, configure the parameters of the OCR engine, including setting the language to Chinese and English, the recognition mode to the default mode, and the confidence interval to 90; S42, text recognition, input the nameplate image to be recognized, and call the OCR engine to complete image recognition and text content extraction; S43, text matching, based on the nameplate template and the text content obtained by OCR engine recognition, the text is segmented and divided, and different texts are divided into different types according to the corresponding positions; according to the template layout and image content, the text content on the image is divided into different types of text boxes; based on the comparison and judgment of the text position on the template and the text position on the collected image, the text content recognized on the nameplate is divided into several different types of text boxes.

7. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Step S5 specifically includes: S51, text classification, classifying the text according to the recognized text content according to the order of system output and the marked text category, and associating it with the corresponding server ledger content in the secondary equipment database of the control cloud system; S52, data default prompt, when the identified equipment parameter does not exist in the ledger information of the corresponding equipment in the secondary equipment database of the control cloud system, mark and prompt the data; when the equipment parameter exists in the secondary equipment database of the control cloud system, but the system does not recognize it, mark and prompt the parameter; when the same field exists in the database and the nameplate identification content, but the corresponding content of the field is empty, mark and prompt the parameter; S53, numerical comparison, when the parameter content on the nameplate of the identification equipment is a number, it is compared with the corresponding content in the secondary equipment database of the control cloud system. When the values ​​of the two are inconsistent, the data is marked; S54, text comparison, when the parameter content on the equipment nameplate is identified as text, NLP technology is used to analyze the text content. When the text content is inconsistent with the secondary equipment database of the control cloud system, the data is marked and prompted.

8. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Step S6 specifically includes: S61, automatic filling of default data. When the data exists in the secondary equipment database of the control cloud system but the system does not recognize the corresponding data, the relevant personnel will be prompted; when the data does not exist in the secondary equipment database of the control cloud system but the system recognizes the corresponding data, the system automatically completes the filling in the database according to the recognition content; S62, value modification, when the identified device parameter is inconsistent with the device parameter in the secondary device database of the control cloud system and the parameter is a numerical type, the system automatically modifies the parameter information according to the identified content and automatically records the modified content; S63, text modification, when the parameter content on the equipment nameplate is identified as text and is inconsistent with the corresponding content in the secondary equipment database of the control cloud system, the text in the database is replaced according to the identified text content, and the modified content is automatically recorded.

9. The method for identifying and verifying nameplate information of electric power equipment based on the control cloud according to claim 1, characterized in that: Also includes: Step S7, the control personnel sets the trigger. When the trigger meets the set time and event requirements, the system automatically verifies the secondary equipment ledger information in the control cloud system, automatically completes the related functions of steps S5 and S6, and stores the verification results and updated content in the system log. The trigger includes a time trigger and a time trigger; The control personnel set the time trigger according to the production demand to realize the automatic inspection of the system. In the time trigger, the control personnel set the number of automatic inspections per day and the time of each inspection. When the system time reaches the set time, the system automatically starts the inspection; The control personnel set event triggers according to production needs to realize automatic inspection of the system. The trigger events set by the event trigger include: equipment inspection timeout, inspection error number exceeds the threshold, equipment ledger information update is completed. When the set trigger event occurs, the event trigger is automatically triggered to carry out automatic inspection of the system again.

10. A power equipment nameplate information recognition and verification system based on control cloud, characterized in that The method applied to any one of claims 1 to 9, wherein the system comprises: The image acquisition module is used to select the secondary equipment in the secondary equipment maintenance management system of the control cloud, take the corresponding nameplate image according to the plant station and equipment name of the secondary equipment, and upload the taken nameplate image to the corresponding database of the control cloud; Template definition module, used to design nameplate templates of relevant nameplate images according to the layout and text content of secondary equipment nameplates; An image preprocessing module is used to preprocess the collected nameplate image to obtain a preprocessed image; The OCR recognition module is used to input the pre-processed image into the OCR processing interface, select the OCR engine, configure relevant parameters and variables, complete the recognition of text content, and classify the recognized text content according to the nameplate template; The ledger comparison module is used to compare the images identified and classified by the OCR engine with the secondary equipment ledger information in the control cloud system, and find and mark the similarities and differences between the two; The ledger modification module is used to modify the corresponding data in the control cloud system based on the ledger errors or omissions found by the system after completing the ledger comparison, and record and prompt the operation and maintenance personnel to modify the content.

Images