Power distribution automation self-service point alignment method based on image text information recognition

Abstract

The invention relates to a power distribution automation self-service point alignment method based on image text information recognition, which comprises the following steps of: 1, shooting or photographing a workstation screen with a power distribution automation system alarm window by using a computer camera, and storing to realize monitoring interface image extraction; 2, performing character information recognition on the alarm window picture obtained in the step 1 by adopting a character recognition model, and formatting mark characters in the character information to enable the information to be accurately divided into rows; 3, extracting telecommand or telemetering change information caused by transmission items line by line from the text lines formatted in the step 2, extracting transmission key information, and structuring the transmission key information; and 4, constructing various transmission logic templates according to terminal types, prompting terminal transmission personnel to carry out transmission by voice item by item based on a transmission item sequence, generating and integrating single-item transmission results, and generating a remote signaling transmission result and a telemetering transmission result of the power distribution terminal. According to the invention, the manpower cost of main station side transmission is greatly reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of power distribution automation, and relates to a remote signaling and remote measurement transmission method of a power distribution automation system, and specifically relates to a self-service point-to-point method for power distribution automation based on image and text information recognition. Background Art

[0002] The construction of new power systems requires continuous improvement of the transparency of distribution networks and the observability and measurability of distribution networks. In recent years, the company has vigorously promoted the construction of distribution automation, and a large number of distribution automation terminals connected to the power system need to carry out transmission work. However, the transmission work requires the operation and maintenance personnel of the distribution automation master station to cooperate with the terminal manufacturers to complete point-by-point verification. Since there are often multiple terminal manufacturers, they hope to carry out transmission work at the same time, but the transmission personnel of the distribution master station are often only 1 to 2, which cannot meet the needs of multiple manufacturers for simultaneous transmission.

[0003] With the development of technologies such as image recognition and voice recognition, there are already some convenient machines in society, such as automatic replacement of driver's licenses, self-service book borrowing and returning, self-service vending machines, and self-service passport replacement. However, the power grid transmission work still relies on the manual monitoring of the alarm window and the front data interface of the master station, which obviously lags behind the development of today's advanced technology. At present, there have been some studies and applications on automatic point-to-point at the substation end and the distribution terminal side. However, the existing research is mostly focused on the substation side and the distribution terminal side, and most of them directly obtain data from the front application of the master station. It is impossible to monitor the final master station alarm window interface, and it is impossible to check the possible faults between the front application of the master station side and the SCADA application in the transmission. Therefore, it is impossible to fundamentally replace the master station transmission personnel. At present, this type of automatic point-to-point device on the market mainly improves the work efficiency of the transmission personnel at the station end, and does not solve the problem of low transmission work efficiency caused by the lack of manpower at the distribution master station.

[0004] In summary, it is necessary to carry out research on self-service point-to-point technology for distribution automation terminals based on image recognition. Image recognition can be used to monitor the alarm window and the front monitoring interface in real time, identify the text and data information in it, and make a judgment on whether the information point transmission has passed based on the information, and feedback to the terminal manufacturer and terminal transmission personnel to replace the monitoring work of the main station personnel, realize the self-service point-to-point of the terminal transmission personnel, and improve the transmission efficiency. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention proposes a distribution automation self-service point-to-point method based on image and text information recognition.

[0006] The above-mentioned object of the present invention is achieved by the following technical solutions:

[0007] A self-service point-to-point method for distribution automation based on image and text information recognition, characterized in that it includes the following steps:

[0008] Step 1: After the self-powered drive starts, use a computer camera with camera and photo-taking functions to capture or take a photo of the workstation screen with the alarm window of the distribution automation system, and store it in the corresponding position on the disk to achieve the extraction of the monitoring interface image;

[0009] Step 2: Adopt an optical character recognition model to recognize the text information of the alarm window picture of the distribution automation system obtained by the camera, convert the picture-based data information into editable text information, and format the flag characters in the text information according to the characteristics of the distribution automation alarm information to accurately separate the information by line;

[0010] Step 3: For the text lines after the preliminary formatting in Step 2, extract the remote signal or remote measurement change information caused by the drive project line by line, extract the feeder, drive terminal name, drive project, and key matters of the alarm content, extract the time text of the alarm information, and convert it into a time format; realize the structuring of the key drive information;

[0011] Step 4: Construct various drive logic templates according to the terminal type, and each type includes two types: the remote signal drive logic template and the remote measurement drive logic template; based on the drive project sequence in the remote signal and remote measurement drive logic templates, prompt the terminal drive personnel to perform the drive item by item, and judge and generate the single-item drive result according to the alarm window text recognition information in Steps 1 and 2 and the drive logic template; give the drive result as a voice feedback to the terminal drive personnel, and receive the voice feedback from the terminal drive personnel to judge and process whether to retransmit; integrate the drive results of a single drive item to automatically generate the drive result of a single information point; integrate the drive results of a single information point to generate the remote signal drive result and remote measurement drive result of the distribution terminal.

[0012] Moreover, in Step 1, the method for image extraction is as follows:

[0013] Call the openCV library of python to read the external camera image, and judge whether the image is successfully read. If the reading is not successful, check the reading program code and the camera opening permission, and then read again. If the reading is successful, obtain the current time, convert the current time into a string format, name the image with the current time, and then save the picture to the specified path to complete the image extraction.

[0014] Moreover, in Step 2, the formatting process is as follows:

[0015] Connect the text list after image recognition into a string with space characters, and process the keywords in the string according to the characteristics of the distribution automation alarm information to form a unified separator - "confirm":

[0016] 1. Replace the characters "Unconfirmed", "Unrestored", and "Confirmed" with "\nConfirmed".

[0017] 2. Replace the character "second" with "second";

[0018] 3. Replace the character "*" with an empty character "".

[0019] 4. Replace the space character " " with an empty character "",

[0020] 5. Replace "second)" with "second) Confirmed"

[0021] 6. Replace "fixed)" with "fixed) Confirmed",

[0022] 7. Replace "Confirmed Confirmed" with "Confirmed".

[0023] Moreover, step 3 includes the extraction and structuring of telemetry key information and the extraction and structuring of telecontrol key information. The processing methods for both are the same. The differences are as follows: For the extraction of telemetry key information, it is also necessary to extract the telemetry value from the alarm information, including:

[0024] 1) First, traverse each line for field structuring, including (1) processing the character "second", replacing "second" and the subsequent blank characters and numbers with "second"; (2) according to the characteristics of distribution automation alarm information, extracting the feeder name, drive interval, alarm content, telemetry value (only required for the extraction of telemetry key information), and information reception time;

[0025] 2) Determine whether the feeder name is extracted. If it is, store the extracted feeder name, drive interval, alarm content, telemetry value (only required for the extraction of telemetry key information), and information reception time in a list, and sort the extracted list rows in ascending order according to the feeder name and drive interval name; if not, determine whether the substation name is extracted. If the substation name is extracted, store the extracted substation name, alarm content, and information reception time in a list, and sort the extracted list rows in ascending order according to the substation name. If the substation name is not extracted, increment the lost information count by 1;

[0026] 3) Save the sorted list as a.csv format file and a.pkl format file to end the process of extracting and structuring telemetry or telecontrol key information.

[0027] Moreover, in step 4, the method for generating the telecontrol drive result and telemetry drive result of the distribution terminal based on the drive logic template is as follows:

[0028] 1) Perform preparatory work: Construct different single-interval drive logic templates according to different terminal types and template types;

[0029] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal drive serial number from the front-end interaction interface;

[0030] 3) Select the single-interval drive logic template according to different terminal types and template types, calculate the single-interval drive point number and the total drive point number. Among them, the single-interval drive point number includes the remote signal single-interval drive point number and the remote measurement single-interval drive point number;

[0031] 4) Delete the existing folders of this terminal and recreate the folders;

[0032] 5) Conduct remote signal drive and result evaluation;

[0033] 6) Conduct remote measurement drive and result evaluation;

[0034] 7) Return the drive terminal serial number, terminal name, number of intervals, total drive point number, and the number of drive points that do not pass; end the drive result evaluation.

[0035] Moreover, in step 4, the method for remote signal drive result evaluation is as follows:

[0036] 1) Read the remote signal part of the single-interval drive logic template;

[0037] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal drive serial number from the front-end interaction interface;

[0038] 3) Select the single-interval drive logic template according to different terminal types and template types, calculate the single-interval drive point number and the total drive point number. Among them, the single-interval drive point number includes the drive point numbers of remote signals and remote measurements;

[0039] 4) Draw the initial progress bar;

[0040] 5) Traverse each switch interval j, traverse each drive item i in the single-interval remote signal drive template, and perform subsequent operations;

[0041] 6) Add the current drive item to the progress set;

[0042] 7) Call the single-item drive result evaluation function, return the single-conversion drive result of the current drive item, and at the same time return the judgment basis;

[0043] 8) Judge whether there is an abnormality. If there is an abnormality, ask whether to redo through voice. If it is necessary to redo, return to step 7). If not, record the drive result and the judgment basis, modify the current drive strategy number i to i + 1, and update the progress bar according to the number of items already added in the progress set. If there is no abnormality, judge whether the drive result passes. If the judgment is yes, enter and update the current drive strategy number i to i + 1 and update the progress bar. If the judgment is no, continue to ask whether to redo through voice;

[0044] 9) Determine whether the current transmission strategy number is greater than the number of transmission items. If it is, generate a telemetry point number based on the switch interval j and the number of transmission items, and modify j to j + 1; if not, return to step 6).

[0045] 10) Determine whether the current switch interval is greater than the number of switch intervals. If it is, complete the evaluation of the telemetry transmission result; if not, return to step 5).

[0046] Moreover, in step 4, the method for telemetry transmission and result judgment is as follows:

[0047] 1) Read the telemetry part of the single-interval transmission logic template.

[0048] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal transmission serial number from the foreground interaction interface.

[0049] 3) Continue to draw the telemetry progress bar based on the telemetry progress bar.

[0050] 4) Traverse each switch interval j and each transmission item i in the single-interval telemetry transmission template, and perform subsequent operations.

[0051] 5) Add the current transmission item to the progress set.

[0052] 6) Call the single-item transmission result evaluation function, return the transmission result of the current telemetry transmission item, and note the telemetry value and whether it exceeds the limit.

[0053] 7) Determine whether there is an abnormality. If there is, ask via voice whether to redo. If yes, return to step 6). If not, record the transmission result and note information, modify the current transmission strategy number i to i + 1, and update the progress bar according to the number of items already added in the progress set. If there is no abnormality, judge whether the transmission result passes. If the judgment is yes, enter and update the current transmission strategy number to i + 1. If the judgment is no, continue to ask via voice whether to redo.

[0054] 8) Determine whether the current transmission strategy number is greater than the number of transmission items. If it is, generate a telemetry point number based on the switch interval j and the number of transmission items, and modify j to j + 1; if not, return to step 5).

[0055] 9) Determine whether the current switch interval is greater than the number of switch intervals. If it is, complete the evaluation of the telemetry transmission result; if not, return to step 4).

[0056] Moreover, in step 4, the method for single-item transmission result evaluation is as follows:

[0057] 1) Obtain the transmission strategy template, interval name, terminal name, and telemetry / telecommand flag from the foreground interaction interface.

[0058] 2) Create a single-project drive strategy folder. If it exists, clear and rebuild it.

[0059] 3) Prompt the manufacturer to start driving the current strategy point via voice. After waiting for a period of time, ask if it has been completed. If not, loop and wait until a completion response is received.

[0060] 4) Take a photo of the alarm window by means of image extraction, save it to the specified path, and display the picture on the foreground GUI interface.

[0061] 5) Start verifying the drive result via voice prompt.

[0062] 6) Filter out the key information lines that are after the drive start time and belong to this terminal.

[0063] 7) Determine whether this project is a telemetry drive project. If it is judged to be yes, enter the subsequent process 7a); if it is judged to be no, enter the subsequent process 7b).

[0064] 7a) The processing procedure is as follows:

[0065] 7a1) Extract key telemetry information.

[0066] 7a2) Calculate the upper and lower limits of telemetry based on the expected transmitted telemetry values and the allowable maximum telemetry error in the telemetry drive template.

[0067] 7a3) Confirm whether the terminal and the drive project are in the traversed information lines extracted by traversing. If so, determine whether the telemetry value is between the upper and lower limits of telemetry. If it is between the limits, notify the site that the drive has passed via voice, return the pass result, and record the telemetry value. If it is not between the limits, determine whether the telemetry value exceeds the upper limit or the lower limit. If it exceeds the upper limit, notify the site that the drive has failed via voice, return the failure result, record the telemetry value and the fact that it exceeds the upper limit. If it exceeds the lower limit, notify the site that the drive has failed via voice, return the failure result, record the telemetry value and the fact that it exceeds the lower limit.

[0068] Confirm whether the terminal and the drive project are in the traversed information lines extracted by traversing. If not, return the failure result and record that the signal is not on.

[0069] 7b) The processing procedure is as follows:

[0070] 7b1) Extract key telemetry signal information.

[0071] 7b2) Count the number of times the main and secondary drive information appears in the photo result after driving.

[0072] 7b3) Determine whether the transmission is passed based on the number of signals. If there is no secondary transmission information for this transmission item in the transmission logic template, then the appearance of the primary transmission information in the transmission result indicates passing; if there is secondary transmission information for this transmission item in the transmission logic template, then the appearance of the primary transmission information in the transmission result and the appearance of all secondary transmission information in the transmission result indicate passing;

[0073] 7b4) Return the result of passing or not passing, and record the occurrence times of the primary and secondary signals as the judgment basis.

[0074] The advantages and positive effects of the present invention are as follows:

[0075] 1. The present invention proposes a method for extracting transmission information based on image text information recognition. Directly read text information from the workstation screen capture based on a trainable text recognition model, and extract transmission information based on picture text recognition. This method does not need to be physically connected to the dispatching technical support system, realizes non-intrusive information acquisition, and greatly reduces the security risk. This method can be applied to the extraction of various system information without being restricted by the availability of interface development.

[0076] 2. The present invention proposes a method for judging the transmission result based on a transmission logic template. Under the unified transmission logic template format, different transmission logic templates can be constructed according to different terminal types. The transmission logic template only needs to contain single-interval transmission item information, and there is no need to input the entire point table. The program can automatically generate telemetry and telecontrol point tables according to the number of intervals input, and sequentially carry out telemetry, telemetry transmission, judge the transmission result, and generate a report record according to the total transmission item order of the user-defined transmission logic template. This method can greatly reduce the labor cost of transmission on the master station side.

[0077] 3. By using the method of the present invention, the distribution automation master station personnel only need to construct the transmission logic template once. After each transmission, only need to input the transmission terminal name, terminal type, number of intervals, select the transmission logic template, and click the start transmission program button to start self-service transmission. It can automatically issue transmission instructions to the substation terminal transmission personnel according to the terminal input order and the order of the transmission logic template, receive the voice response of the substation terminal transmission personnel, and identify it as text, realize the automatic extraction of the alarm window image, text information recognition, key information extraction, transmission result judgment, judgment basis record, voice feedback to the terminal personnel, and generate telemetry and telemetry transmission reports. Thus, the efficiency of inputting the transmission point table in the present invention can be improved. Brief Description of the Drawings

[0078] Figure 1 is the flowchart of the distribution automation self-aligning method based on image text information recognition of the present invention;

[0079] Figure 2It is the decomposition diagram of the camera monitoring interface extraction function module of the present invention;

[0080] Figure 3 It is the flow chart of the camera startup of the present invention;

[0081] Figure 4 It is the flow chart of image extraction of the present invention;

[0082] Figure 5 It is the flow chart of turning off the camera of the present invention;

[0083] Figure 6 It is the flow chart of the image text line-by-line recognition program based on paddleOCR of the present invention;

[0084] Figure 7 It is the classification block diagram of the key information extraction of the transmission of the present invention;

[0085] Figure 8 It is the process of key information extraction and structured processing of telemetry of the present invention;

[0086] Figure 9 It is the process of key information extraction and structured processing of teleindication of the present invention;

[0087] Figure 10 It is the classification block diagram of the transmission result evaluation of the present invention;

[0088] Figure 11 It is the flow chart of the transmission result evaluation of a single power distribution terminal of the present invention;

[0089] Figure 12 It is the flow chart of the transmission result evaluation of teleindication of the present invention;

[0090] Figure 13 It is the flow chart of the transmission result evaluation of telemetry of the present invention;

[0091] Figure 14 It is the flow chart of the transmission result evaluation of a single item of the present invention;

[0092] Figure 15 It is the flow chart of text-to-speech of the present invention;

[0093] Figure 16 It is the flow chart of speech-to-text of the present invention;

[0094] Figure 17 It is the block diagram of the automatic generation type of the transmission report of the present invention;

[0095] Figure 18 It is the program logic diagram of the teleindication report generation of the present invention;

[0096] Figure 19 It is the program logic diagram of the telemetry report generation of the present invention;

[0097] Figure 20 This is a sample interface diagram before the transmission of the present invention starts;

[0098] Figure 21 This is a sample interface diagram after the transmission of the present invention ends;

[0099] Figure 22 This is a sample interface diagram for opening the transmission report (telemetry) of the present invention;

[0100] Figure 23 This is a sample interface diagram for opening the transmission report (tele-signal) of the present invention;

[0101] Figure 24 This is a screenshot of the transmission result report of the tele-signal of the two-section switch station of DLK1941 of the present invention;

[0102] Figure 25 This is a screenshot of the transmission result report of the telemetry of the two-section switch station of DLK1941 of the present invention. Detailed implementation manners

[0103] The structure of the present invention will be further described below with reference to the accompanying drawings and through embodiments. It should be noted that this embodiment is narrative rather than restrictive.

[0104] A distribution automation self-aligning point method based on image text information recognition, please refer to Figures 1 - 25 , including the following steps:

[0105] Step 1: After the self-service transmission starts, extract the monitoring interface image

[0106] Use a computer camera with a camera and photographing function to photograph or take pictures of the workstation screen with a distribution automation system alarm window, and store it at the corresponding position on the disk. This step is mainly to obtain the transmission monitoring interface in the form of pictures, which is convenient for subsequent picture data processing.

[0107] In the present invention, a computer is connected to an external 4K high-definition conference camera. When the program needs to obtain the information of the alarm window, the openCV library of python is called to start, take pictures, and release the resources of the camera, and the photographing result is stored in the corresponding path of the computer for subsequent image recognition programs to identify and analyze. It includes the camera startup process, the image extraction process, and the camera closing process. Please refer to Figures 2 - 5 .

[0108] Among them, the method for image extraction is:

[0109] Call the openCV library of Python to read images from an external camera, and determine whether the image is successfully read. If the reading is not successful, check the program code for reading and the camera access permission, and then perform the reading again. If the reading is successful, obtain the current time, convert the current time to a string format, use the current time as the image name, and then save the picture to the specified path to complete the extraction of the image.

[0110] Step 2: Perform image text information recognition on the images extracted in Step 1

[0111] Adopt open-source text recognition models such as paddleOCR to perform text information recognition on the pictures of the distribution automation system alarm window obtained from the camera, convert the picture-based data information into editable text information, and perform formatting processing on the flag characters in the text information according to the characteristics of the distribution automation alarm information to accurately separate the information into lines for subsequent extraction of key transmission information. The specific formatting processing is as follows:

[0112] Use space characters to concatenate the text list after image recognition into a string, and process the keywords in the string according to the characteristics of the distribution automation alarm information to form a unified delimiter - "confirmation", see Figure 6 :

[0113] 1. Replace the characters "unconfirmed", "unrestored", and "confirmed" with "\nconfirmation"

[0114] 2. Replace the character "seconds" with "seconds";

[0115] 3. Replace the character "*" with an empty character ""

[0116] 4. Replace the space character "" with an empty character "",

[0117] 5. Replace "seconds)" with "seconds) confirmation"

[0118] 6. Replace "fixed)" with "fixed) confirmation",

[0119] 7. Replace "confirmation confirmation" with "confirmation"

[0120] Step 3: Extract key transmission information from the text lines after the preliminary formatting processing in Step 2

[0121] For the text lines after the preliminary formatting processing, extract the telemetry or telemetry changes caused by the transmission information points line by line, extract key items such as the feeder, transmission terminal name, transmission item, and alarm content, extract the time text of the alarm information, and convert it into a time format; realize the structuring of the key transmission information. This includes the extraction and structuring of key telemetry information and the extraction and structuring of key telemetry information, see Figures 7 - 9 。

[0122] The method for extracting and structuring key telemetry information is as follows:

[0123] 1) First, traverse each line and perform field structuring, including (1) processing the "second" character, replacing the "second" and the subsequent whitespace characters and numbers with "second"; (2) according to the characteristics of distribution automation alarm information, extract the feeder name, drive interval, alarm content, telemetry value, and information reception time;

[0124] 2) Determine whether the feeder name is extracted. If it is, store the extracted feeder name, drive interval, alarm content, telemetry value, and information reception time in a list, and sort the extracted list rows in ascending order according to the feeder name and drive interval name; if not, determine whether the substation name is extracted. If the substation name is extracted, store the extracted substation name, alarm content, and information reception time in a list, and sort the extracted list rows in ascending order according to the substation name. If the substation name is not extracted, increment the lost information count by 1;

[0125] 3) Save the sorted list as a.csv format file and a.pkl format file to end the process of extracting and structuring key telemetry information.

[0126] The method for extracting and structuring key teleindication information refers to the above method for extracting and structuring key telemetry information, which will not be elaborated here. The processing methods of the two are the same. The difference is that the extraction of key telemetry information also needs to extract the telemetry value from the alarm information.

[0127] Step 4: Conduct drive result evaluation

[0128] Construct a drive logic template according to the terminal type, and generate the teleindication drive result and telemetry drive result of the distribution terminal based on the drive logic template. Under the unified drive logic template format, different drive logic templates can be constructed according to different terminal types. The drive logic template only needs to contain single-interval drive item information and does not need to enter the entire point table. The program can automatically generate teleindication and telemetry point tables according to the input number of intervals, and sequentially perform teleindication, teleindication drive, drive result evaluation, and generate report records according to the total drive item sequence of the user-defined drive logic template.

[0129] For the evaluation of teleindication drive results, telemetry drive results, and single-item drive results, see Figures 10 - 14 , as shown in Table 1 and Table 2

[0130] Table 1. Teleindication Drive Logic Template (DTU)

[0131]

[0132]

[0133] Table 2. Telemetry Transmission Logic Template (DTU)

[0134]

[0135] The method for generating the transmission result of a single power distribution terminal based on the transmission logic template is as follows:

[0136] 1) Prepare: Construct different single - interval transmission logic templates according to different terminal types and template types;

[0137] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal transmission serial number from the front - end interaction interface;

[0138] 3) Select the single - interval transmission logic template according to different terminal types and template types, calculate the single - interval transmission point number and the total transmission point number, where the single - interval transmission point number includes the telemetry single - interval transmission point number and the telemetry single - interval transmission point number;

[0139] 4) Clear the existing folder of the terminal and recreate the folder;

[0140] 5) Conduct telemetry transmission and result evaluation;

[0141] 6) Conduct telemetry transmission and result evaluation;

[0142] 7) Return the transmission terminal serial number, terminal name, number of intervals, total transmission point number, and the number of transmission points that failed; end the transmission result evaluation.

[0143] The method for evaluating the telemetry transmission result is as follows:

[0144] 1) Read the telemetry part of the single - interval transmission logic template;

[0145] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal transmission serial number from the front - end interaction interface;

[0146] 3) Select the single - interval transmission logic template according to different terminal types and template types, calculate the single - interval transmission point number and the total transmission point number, where the single - interval transmission point number includes the transmission point numbers of telemetry and telemetry;

[0147] 4) Draw the initial progress bar;

[0148] 5) Traverse each switch interval j, traverse each transmission item i in the single - interval telemetry transmission template, and perform subsequent operations;

[0149] 6) Add the current transmission item to the progress set;

[0150] 7) Call the single-item drive result evaluation function to return the single-displacement drive result of the current drive item, and at the same time return the judgment basis;

[0151] 8) Determine whether there is an abnormality. If there is, ask via voice whether to redo. If it is necessary to redo, return to step 7). If not, record the drive result and the judgment basis, modify the current drive strategy number i to i + 1, and update the progress bar according to the number of items already added in the progress set; if there is no abnormality, judge whether the drive result passes. If the judgment is yes, enter the update of the current drive strategy number i to i + 1 and update the progress bar. If the judgment is no, continue to ask via voice whether to redo the judgment;

[0152] 9) Determine whether the current drive strategy number is greater than the number of drive items. If so, generate a telecontrol signal point number according to the switch interval j and the number of drive items, and modify j to j + 1; if not, return to step 6)

[0153] 10) Determine whether the current switch interval is greater than the number of switch intervals. If so, complete the evaluation of the telecontrol drive result. If not, return to step 5).

[0154] The above method for telemetry drive and result judgment is as follows:

[0155] 1) Read the telemetry part of the single-interval drive logic template;

[0156] 2) Obtain the terminal name, number of intervals, terminal type, template type, and terminal drive serial number from the foreground interaction interface;

[0157] 3) Continue to draw the telemetry progress bar based on the telecontrol progress bar;

[0158] 4) Traverse each switch interval j and each drive item i in the single-interval telemetry drive template, and perform subsequent operations;

[0159] 5) Add the current drive item to the progress set;

[0160] 6) Call the single-item drive result evaluation function to return the drive result of the current telemetry drive item, and at the same time note the telemetry value and whether it exceeds the limit;

[0161] 7) Determine whether there is an abnormality. If there is, ask via voice whether to redo. If it is necessary to redo, return to step 6). If not, record the drive result and the note information, modify the current drive strategy number i to i + 1, and update the progress bar according to the number of items already added in the progress set; if there is no abnormality, judge whether the drive result passes. If the judgment is yes, enter the update of the current drive strategy number to i + 1. If the judgment is no, continue to ask via voice whether to redo the judgment;

[0162] 8) Determine whether the current transmission strategy number is greater than the number of transmission items. If it is, generate a telemetry point number based on the switch interval j and the number of transmission items, and modify j to j + 1. If not, return to step 5).

[0163] 9) Determine whether the current switch interval is greater than the number of switch intervals. If it is, complete the evaluation of the telemetry transmission result. If not, return to step 4).

[0164] The method for evaluating the transmission result of a single item is as follows:

[0165] 1) Obtain the transmission strategy template, interval name, terminal name, and telemetry / telecommand flag from the foreground interaction interface.

[0166] 2) Create a single-item transmission strategy folder. If it exists, clear and rebuild it.

[0167] 3) Use voice prompts to notify the manufacturer to start transmitting the current strategy point. After waiting for a period of time, ask if it has been completed. If not, loop and wait until a completion response is received.

[0168] 4) Take a photo of the alarm window through image extraction means, save it to the specified path, and display the picture on the foreground GUI interface.

[0169] 5) Start verifying the transmission result through voice prompts.

[0170] 6) Filter out the key information lines after the transmission start time and belonging to the terminal.

[0171] 7) Determine whether the item is a telemetry transmission item. If it is judged to be yes, enter the subsequent processing process 7a). If it is judged to be no, enter the subsequent processing process 7b).

[0172] The processing process of 7a) is as follows:

[0173] 7a1) Extract key telemetry information.

[0174] 7a2) Calculate the upper and lower limits of telemetry based on the expected transmitted telemetry value and the allowable maximum telemetry error in the telemetry transmission template.

[0175] 7a3) Confirm whether the terminal and transmission item are in the traversed key information lines. If so, determine whether the telemetry value is between the upper and lower limits of telemetry. If it is between the limits, notify the site that the transmission is passed by voice, return the passing result, and record the telemetry value. If it is not between the limits, determine whether the telemetry value exceeds the upper limit or the lower limit. If it exceeds the upper limit, notify the site that the transmission fails by voice, return the failing result, and record the telemetry value and the fact that it exceeds the upper limit. If it exceeds the lower limit, notify the site that the transmission fails by voice, return the failing result, and record the telemetry value and the fact that it exceeds the lower limit.

[0176] Confirm whether the terminal and transmission items are in the traversed key information lines extracted by traversal. If not, return a non-pass result and record the absence of a signal.

[0177] 7b) The processing procedure is as follows:

[0178] 7b1) Extract key information of tele-signals.

[0179] 7b2) Count the number of times the main and secondary transmission information appears in the photographed results after transmission.

[0180] 7b3) Determine whether the transmission is passed based on the number of signals. If there is no secondary transmission information for this transmission item in the transmission logic template, the appearance of the main transmission information in the transmission result means passing; if there is secondary transmission information for this transmission item in the transmission logic template, the appearance of the main transmission information in the transmission result and the appearance of all secondary transmission information in the transmission result mean passing.

[0181] 7b4) Return the result of passing or not passing, and record the number of appearances of the main and secondary signals as the basis for judgment.

[0182] Conduct transmission voice interaction and report generation:

[0183] Feed back the transmission result to the terminal transmission personnel. If it passes, give a voice feedback of passing the verification to the terminal manufacturer and the terminal transmission personnel, and continue to carry out the verification of new transmission information in the follow-up; if it does not pass, give a voice feedback of the transmission failure result to the terminal transmission personnel, and ask whether to re-transmit. If the on-site terminal transmission personnel need to re-transmit, wait for a certain period of time, re-conduct voice interaction and result verification. If they do not need to re-transmit, continue to carry out the voice prompt for the next point of transmission. Automatically record information such as transmission time, transmission process, transmission result, and transmission problems, and generate a visual report. This includes issuing voice commands, receiving voice information, and automatically generating a transmission report. See Figures 15 - 19 .

[0184] Issue a voice command to achieve text-to-speech conversion. The conversion process is as follows: Initialize the speech synthesis engine; set the speech rate and volume parameters; save the audio file and play the audio to achieve text-to-speech conversion.

[0185] Receive voice information to achieve speech-to-text conversion. The conversion process is as follows: Start the speech recognition engine; load the keyword dictionary; conduct real-time speech collection and recognition; conduct keyword matching judgment. If the match is successful, stop the collection, save the result data, and achieve speech-to-text conversion; if the match fails, give a voice prompt to retry, and then conduct real-time speech collection and recognition.

[0186] Automatically generating a transmission report includes generating a tele-signal transmission report and a telemetry transmission report.

[0187] The generation process of the remote signal transmission report is as follows:

[0188] 1) Extract the transmission items belonging to the unified point numbers in the remote signal transmission results, comprehensively judge the transmission results. If all are passed, the transmission result of this information point is passed; otherwise, it is not passed. Convert the transmission result to the point table mode.

[0189] 2) Count the number of remote signal points that fail to pass for this terminal.

[0190] 3) Generate the transmission results into an Excel transmission report, and adjust the table formats such as font and line spacing to complete the generation of the remote signal transmission report.

[0191] The generation process of the remote measurement transmission report is as follows:

[0192] 1) Count the number of remote measurement points that fail to pass for this terminal.

[0193] 2) Generate the transmission results into the Excel transmission report and adjust the table formats such as font and line spacing.

[0194] 3) Save the generated remote measurement transmission report to the result folder to complete the generation of the remote measurement transmission report.

[0195] The method of the present invention is completed under the human-computer interaction interface, which is the transmission interface of the distribution automation self-aligning point system, mainly including three parts: information entry before transmission, transmission process display, and transmission result query.

[0196] Before the transmission starts, the main station operation and maintenance personnel need to enter the name of the distribution equipment terminal, the number of switch intervals, the terminal type, and the template selection information for this transmission. Click "Confirm Entry", then all the entered terminal information will be displayed in the text box at the lower left, and the number of entered terminals will also be displayed. Click "Re-enter", then the entered information will be cleared. After confirming that the entered information is correct, click "Start Transmission", then the program will start to carry out the transmission work item by item according to the entered terminal order and the transmission logic template. At the beginning of the transmission, the program asks whether to adjust the camera position. The user can adjust the camera angle according to the camera-taken pictures displayed on the right until all the warning messages are displayed in the camera-taken pictures and try to keep the text horizontal as much as possible.

[0197] During the transmission process, the time consumed for the left-side transmission can be counted as the time from the start of the click for transmission to the current transmission item; the progress bar in the middle of the transmission process shows the proportion of the currently transmitted items to the total number of transmitted items; the text box in the upper middle shows various information during the transmission process, including the voice command text sent, the transmission process information, the voice recognition text, and the intermediate transmission results; the updated right side shows the picture of the warning window captured by the camera; the list in the lower middle shows the terminal information that has been transmitted, including the name, interval number, type, total number of transmission points, and number of unqualified points. During the transmission process, the program automatically sends voice commands and receives voice information to judge and record the transmission results without manual intervention.

[0198] After the transmission ends, the list in the lower middle shows all the terminal information that has been transmitted, including the name, interval number, type, total number of transmission points, and number of unqualified points. Click to select a terminal, and you can select the name of the terminal for which you need to preview the report. Click to export the report, and you can view the telecommunication transmission report and the telemetry transmission report of this terminal.

[0199] Click the x in the upper right corner to close the program.

[0200] An example of the program's human-computer interaction interface is as follows:

[0201] Before the transmission starts, see Figure 20 ; after the transmission ends, see Figure 21 ; an example diagram of the opening interface of the transmission report (telemetry), see Figure 22 ; an example diagram of the opening interface of the transmission report (telecommunication), see Figure 23 ; a screenshot of the telecommunication transmission result report of the two-interval switch station of DLK1941, see Figure 24 ; a screenshot of the telemetry transmission result report of the two-interval switch station of DLK1941, see Figure 25 .

[0202] Although the embodiments and drawings of the present invention are disclosed for illustrative purposes, those skilled in the art can understand that various substitutions, changes, and modifications are possible without departing from the spirit of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the content disclosed in the embodiments and drawings.

Claims

1. A self-service point-to-point method for distribution automation based on image and text information recognition, characterized in that: The steps are as follows: Step 1: After the self-powered drive starts, use a computer camera with a camera and photo-taking function to take a video or photo of the workstation screen with a distribution automation system alarm window, and store it at the corresponding position on the disk to realize the extraction of the monitoring interface image; Step 2: Adopt an optical character recognition model to recognize the text information of the distribution automation system alarm window picture obtained by the camera, convert the picture data information into editable text information, and format the flag characters in the text information according to the characteristics of the distribution automation alarm information to accurately separate the information by lines; Step 3: For the text lines after the preliminary formatting process in Step 2, extract the telemetry or telemetry change information caused by the drive item line by line, extract the feeder, drive terminal name, drive item, and key matters of the alarm content, extract the time text of the alarm information, and convert it into a time format; realize the structuring of the drive key information; Step 4: Construct various drive logic templates according to the terminal type, each of which includes two types: the telemetry drive logic template and the telemetry drive logic template; based on the drive item order in the telemetry and telemetry drive logic templates, prompt the terminal drive personnel to perform the drive item by item, and judge and generate the single-item drive result according to the alarm window text recognition information and the drive logic template in Steps 1 and 2; give the voice feedback of the drive result to the terminal drive personnel, and receive the voice feedback of the terminal drive personnel to judge and process whether to retransmit; integrate the drive results of a single drive item to automatically generate the drive result of a single information point; integrate the drive results of a single information point to generate the telemetry drive result and the telemetry drive result of the distribution terminal.

2. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In Step 1, the method for image extraction is as follows: Call the openCV library of python to read the external camera image, and judge whether the image is successfully read. If the reading is not successful, check the reading program code and the camera opening permission, and then read again. If the reading is successful, obtain the current time, convert the current time into a string format, use the current time as the image name, and then save the picture to the specified path to complete the image extraction.

3. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In Step 2, the formatting process is as follows: Use space characters to connect the text list after image recognition into a string, and process the keywords in the string according to the characteristics of the distribution automation alarm information to form a unified separator - "confirmed":

1. Replace the characters "unconfirmed", "unrestored", and "confirmed" with "\nconfirmed" 2. Replace the character "seconds" with "seconds"; 3. Replace the character "*" with an empty character "" 4. Replace the space character "" with an empty character "" 5. Replace "seconds)" with "seconds) confirmed" 6. Replace "fixed)" with "fixed) confirmed" 7. Replace "confirmed confirmed" with "confirmed".

4. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: Step 3 includes the extraction and structuring of telemetry key information and the extraction and structuring of telemetry key information. The processing methods of both are the same, including: 1) First, traverse each line and perform field structuring, including (1) processing the "second" character and replacing the "second" and the blank characters and numbers after it with "second"; (2) extracting the feeder name, transmission interval, alarm content, telemetry value, and information reception time according to the characteristics of distribution automation alarm information; 2) Determine whether the feeder name is extracted. If so, store the extracted feeder name, transmission interval, alarm content, telemetry value and information receiving time in a list, and sort the extracted list rows in ascending order according to the feeder name and transmission interval name; if not, determine whether the substation name is extracted. If the substation name is extracted, store the extracted substation name, alarm content and information receiving time in a list, and sort the extracted list rows in ascending order according to the substation name. If the substation name is not extracted, add 1 to the missing information count; 3) Save the sorted list in step 3 as a .csv format file and a .pkl format file, completing the telemetry or telesignaling key information extraction and structuring process.

5. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In step 4, the method for generating the remote signal transmission result and the remote measurement transmission result of the distribution terminal based on the transmission logic template is: 1) Perform preparation work: construct different single-bay transmission logic templates according to different terminal types and template types; 2) Obtain the terminal name, interval number, terminal type, template type, and terminal transmission serial number from the front-end interactive interface; 3) Select the single-interval transmission logic template according to different terminal types and template types, and calculate the single-interval transmission points and the total transmission points. The single-interval transmission points include the telesignaling single-interval transmission points and the telemetry single-interval transmission points. 4) Clear the existing folders on the terminal and create new folders; 5) Carry out remote signal transmission and result evaluation; 6) Conduct telemetry transmission and evaluate the results; 7) Return the transmission terminal serial number, terminal name, number of intervals, total transmission points and number of failed transmission points; end the transmission result evaluation.

6. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In step 4, the method for judging the remote signal transmission result is: 1) Read the remote signal part of the single-bay transmission logic template; 2) Obtain the terminal name, interval number, terminal type, template type, and terminal transmission serial number from the front-end interactive interface; 3) Select the single-interval transmission logic template according to different terminal types and template types, and calculate the single-interval transmission points and the total transmission points, where the single-interval transmission points include the transmission points of telesignaling and telemetry; 4) Draw the initial progress bar; 5) Traverse each switch bay j, traverse each transmission item i in the single bay remote signal transmission template, and perform subsequent operations; 6) The current transmission project is added to the progress collection; 7) Call the single-item transmission result judgment function to return the single-displacement transmission result of the current transmission item and the judgment basis; 8) Determine whether there is an abnormality. If so, ask whether it needs to be redone by voice. If it needs to be redone, return to step 7). If not, record the transmission result and the judgment basis, and modify the current transmission strategy number i to i+1, and update the progress bar according to the number of added items in the progress set; if there is no abnormality, determine whether the transmission result is passed. If it is yes, update the current transmission strategy number i to i+1 and update the progress bar. If it is no, continue to ask by voice whether it needs to be redone; 9) Determine whether the current transmission strategy number is greater than the number of transmission items. If so, generate the remote signal point number according to the switch interval j and the number of transmission items, and modify j to j+1; if not, return to step 6) 10) Determine whether the current switch interval is greater than the number of switch intervals. If so, complete the remote signal transmission result evaluation. If not, return to step 5).

7. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In step 4, the method of telemetry transmission and result judgment is: 1) Read the telemetry part of the single-interval transmission logic template; 2) Obtain the terminal name, interval number, terminal type, template type and terminal transmission serial number from the front-end interactive section; 3) Continue to draw the telemetry progress bar based on the telesignaling progress bar; 4) Traverse each switch bay j, traverse each transmission item i in the single bay telemetry transmission template, and perform subsequent operations; 5) Add the current transmission project to the progress set; 6) Call the single-item transmission result evaluation function to return the transmission result of the current telemetry transmission item, and note the telemetry value and whether it exceeds the limit; 7) Determine whether there is an abnormality. If so, ask whether it needs to be redone through voice. If it needs to be redone, return to step 6). If not, record the transmission result and remarks, and modify the current transmission strategy number i to i+1, and update the progress bar according to the number of items added in the progress set; If there is no abnormality, the transmission result is judged whether it passes. If it is judged to be yes, the current transmission strategy number is updated to i+1. If it is judged to be no, the voice inquiry is continued to whether the judgment needs to be redone; 8) Determine whether the current transmission strategy number is greater than the number of transmission items. If so, generate the telemetry point number based on the switch interval j and the number of transmission items, and modify j to j+1; if not, return to step 5) 9) Determine whether the current switch interval is greater than the number of switch intervals. If so, complete the telemetry transmission result evaluation. If not, return to step 4).

8. The self-service point-to-point method for distribution automation based on image and text information recognition according to claim 1 is characterized in that: In step 4, the method for judging the transmission results of a single project is: 1) Obtain the transmission strategy template, interval name, terminal name, and telemetry and telesignaling flag from the front-end interactive interface; 2) Create a single-project transmission strategy folder. If it already exists, clear and rebuild it. 3) The voice prompts the manufacturer to start the transmission of the current strategic point. After waiting for a while, it asks whether it has been completed. If it has not been completed, it will wait in a loop until it receives a reply that it has been completed; 4) Take a picture of the alarm window by image extraction, save it to the specified path, and display the picture on the front-end GUI interface; 5) Start verifying the transmission results through voice prompts; 6) Filter out the key information rows that are after the transmission start time and belong to the terminal; 7) Determine whether the project is a telemetry transmission project. If it is judged to be yes, then enter the subsequent 7a) processing process; if it is judged to be no, then enter the subsequent 7b) processing process; 7a) The processing process is: 7a1) Extracting key telemetry information; 7a2) Calculate the upper and lower limits of telemetry according to the telemetry value to be sent in the telemetry transmission template and the maximum allowable telemetry error; 7a3) Confirm whether the terminal and transmission items are in the traversed row after traversing the key information extracted. If so, determine whether the telemetry value is between the upper and lower limits of telemetry. If it is between the upper and lower limits, voice notify the site that the transmission is passed, and return the pass result and record the telemetry value. If it is not between the upper and lower limits, determine whether the telemetry value exceeds the upper limit or the lower limit. If it exceeds the upper limit, voice notify the site that the transmission fails, return the fail result and record the telemetry value and the upper limit. If it exceeds the lower limit, voice notify the site that the transmission fails, return the fail result and record the telemetry value and the lower limit. Confirm the key information rows extracted by traversal to determine whether the terminal and transmission items are in the traversal rows. If not, return a failed result and record a failed signal; 7b) The processing process is: 7b1) Extracting key information of remote signaling; 7b2) Count the number of times the primary and secondary transmission information appears in the photo results after the transmission; 7b3) Determine whether the transmission is passed according to the number of signals. If the transmission item in the transmission logic template does not have secondary transmission information, the transmission is passed if the primary transmission information appears in the transmission result; if the transmission item in the transmission logic template has secondary transmission information, the transmission is passed if the primary transmission information appears in the transmission result and the secondary transmission information also appears in the transmission result; 7b4) Returns a pass or fail result, recording the number of occurrences of the primary and secondary signals as a basis for judgment.

Images