CN110807460A - A kind of substation intelligent patrol system based on image recognition and its application method - Google Patents

Abstract

The invention discloses a substation intelligent inspection system based on image recognition and an application method thereof. The substation intelligent inspection system includes an online monitoring sub-station, an online monitoring main station, an intelligent operation and maintenance main station, an image recognition server and a mobile operation and maintenance terminal. The monitoring sub-station, the online monitoring master station, and the intelligent operation and maintenance master station are connected in sequence, the intelligent operation and maintenance master station and the image recognition server are connected, and the intelligent operation and maintenance master station and the mobile operation and maintenance terminal are connected through a network. Application method Including the intelligent operation and maintenance main station calling the image recognition server to perform image recognition according to the image collected by the mobile operation and maintenance terminal, and combined with the state monitoring data of the secondary equipment to generate the inspection result and output it to the mobile operation and maintenance terminal. The invention can realize the automatic inspection of the substation, can effectively improve the efficiency and accuracy of the inspection of the substation, reduce the inspection workload of the inspectors, and improve the intelligent inspection level of the substation, thereby effectively ensuring the safe and stable operation of the substation.

Description

A kind of substation intelligent patrol system based on image recognition and its application method

technical field

The invention relates to the technical field of power system operation and maintenance, in particular to a substation intelligent patrol system based on image recognition and an application method thereof.

Background technique

The inspection of the substation is an important work content in the daily operation and maintenance process of the substation. The inspection content of the substation is generally to check whether the equipment status in the substation conforms to the current operating state of the substation. Secondary equipment that does not conform to the current operating state of the substation should be inspected in place in time and dealt with accordingly. At present, the inspection task of the substation is generally carried out manually, and the inspection of the secondary equipment status of the whole station is carried out manually at regular intervals, so as to discover the hidden defects of the substation in time. However, by manually inspecting the status of secondary equipment in substations, there are problems such as long inspection time, difficulty in ensuring the correctness of inspection results relying on manual judgment, and difficulty in ensuring inspection efficiency. In order to improve the inspection efficiency and inspection accuracy of substations, a new operation and maintenance method is urgently needed, which can effectively solve the problems existing in substation inspections, reduce the workload of operation and maintenance personnel, improve the quality of substation inspections, and improve the automation of substation inspections, so as to effectively Ensure the safe and stable operation of the substation.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention: in view of the above problems of the prior art, an image recognition-based intelligent inspection system for substations and an application method thereof are provided. The present invention can realize automatic inspection of substations, and can effectively improve the efficiency and accuracy of inspection of substations. It reduces the inspection workload of the inspectors and improves the intelligent inspection level of the substation, thereby effectively ensuring the safe and stable operation of the substation.

In order to solve the above-mentioned technical problems, the technical scheme adopted in the present invention is:

A substation intelligent inspection system based on image recognition, including online monitoring sub-station, online monitoring main station, intelligent operation and maintenance main station, image recognition server and mobile operation and maintenance terminal, the online monitoring sub-station, online monitoring main station, intelligent operation and maintenance terminal The operation and maintenance master stations are connected in sequence through the network, the intelligent operation and maintenance master station and the image recognition server are connected through the network, the intelligent operation and maintenance master station and the mobile operation and maintenance terminal are connected through the network, and the online monitoring The input end of the sub-station is connected with the monitored secondary equipment to send the state monitoring data of the secondary equipment composed of the collected secondary equipment status and alarm information to the online monitoring master station, and the online monitoring master station sends each online monitoring station to the online monitoring master station. The state monitoring data of the secondary equipment of the sub-station is aggregated and pushed to the main intelligent operation and maintenance station. The status monitoring data generates inspection results and outputs them to the mobile operation and maintenance terminal.

Preferably, it also includes information security zone I, information security zone II, and information security zone III with a total of three physical space-isolated information security zones, and the online monitoring sub-station and its monitored secondary equipment are located in the information security zone I, The online monitoring master station is located in the information security zone II, and the intelligent operation and maintenance master station, the image recognition server and the mobile operation and maintenance terminal are located in the information security zone III.

Preferably, the network between the online monitoring sub-station, the online monitoring main station, and the intelligent operation and maintenance main station adopts the MMS communication protocol for communication connection, and the input end of the online monitoring sub-station and the monitored secondary equipment adopt MMS Communication protocol for communication connection.

Preferably, the intelligent operation and maintenance master station and the mobile operation and maintenance terminal are connected through a power wireless private network.

Preferably, a private protocol is used for communication between the intelligent operation and maintenance master station and the image recognition server.

In addition, the present invention also provides an application method of the aforementioned image recognition-based intelligent inspection system for substations, and the implementation steps include:

1) Select the target secondary equipment, the intelligent operation and maintenance master station collects the status monitoring data of the target secondary equipment through the online monitoring sub-station and the online monitoring master station, and automatically selects the corresponding secondary equipment state database according to the current secondary equipment operating status , the secondary equipment state library includes the current state of the target secondary equipment; the mobile operation and maintenance terminal uses its own camera to collect on-site inspection images for the target secondary equipment, and upload the inspection images to the intelligent operation and maintenance master through the network stand;

2) The intelligent operation and maintenance master station sends the inspection image to the image recognition server;

3) The image recognition server performs image recognition on the patrol image to obtain the image recognition object status of the recognition secondary equipment, and returns the image recognition result to the intelligent operation and maintenance master station;

4) The intelligent operation and maintenance master station determines the image recognition object status of the target secondary equipment according to the image recognition object status of the recognition secondary equipment obtained by image recognition, and compares the image recognition object status of the target secondary equipment and its corresponding secondary equipment. The state library checks the state information. If the state of the image recognition object is consistent with the state of the secondary equipment state library, the automatic inspection result is determined to be normal, otherwise it is determined to be abnormal;

5) The intelligent operation and maintenance master station will push the automatic inspection results to the mobile operation and maintenance terminal, and the on-site personnel will manually confirm the results of the next secondary equipment inspection or the current secondary equipment inspection according to the automatic inspection results.

Preferably, the step 1) further includes the step of establishing a secondary equipment state library, and the establishment of the secondary equipment state library is specifically one of the following methods ① and ② or a combination of the two methods, and the method ①: For each state of the target secondary device, the secondary device state library of the target secondary device in this state is obtained by manually establishing the secondary device and confirming the state of the secondary device; method ②: for each state of the target secondary device In a state, the on-site secondary equipment images are collected by on-site personnel respectively, and image recognition is performed to automatically enter the secondary equipment and secondary equipment states to obtain the secondary equipment state library of the target secondary equipment in this state.

Preferably, the step of determining the image recognition object state of the target secondary device according to the image recognition object state of the recognition secondary device obtained by image recognition in step 4) includes:

4.1) Obtain the object rectangle for recognizing secondary equipment and the rectangle for recognizing text in the image recognition result;

4.2) Match the rectangular frame of the object for identifying the secondary device and the rectangular frame of the text to obtain the object text for identifying the secondary device;

4.3) Match each recognition secondary device and its object text with the target secondary device to obtain the image recognition result composed of the object rectangular frame and the object text of the target secondary device.

Preferably, the detailed steps of step 4.2) include: traversing the rectangular frame of the recognized text around it according to the coordinates of the object rectangular frame for recognizing the secondary device, and determining the rectangular frame of the recognized text closest to the object rectangular frame for recognizing the secondary device as the recognized text In order to identify the object character of the identification secondary device.

Preferably, the detailed steps of step 4.3) include: matching each recognition secondary device and its object text with the target secondary device, so as to obtain an image recognition result composed of the object rectangular frame and the object text of the target secondary device.

Compared with the prior art, the present invention has the following advantages: the input end of the online monitoring sub-station of the present invention is connected with the monitored secondary equipment to monitor the state of the secondary equipment composed of the collected secondary equipment state and alarm information. The data is sent to the online monitoring master station, and the online monitoring master station aggregates the status monitoring data of the secondary equipment of each online monitoring substation and pushes it to the intelligent operation and maintenance master station, which is used for mobile operation and maintenance. The image collected by the maintenance terminal calls the image recognition server for image recognition and combines the status monitoring data of the secondary equipment to generate the inspection result and output it to the mobile operation and maintenance terminal, so it can realize the automatic inspection of the substation, which can effectively improve the efficiency and accuracy of the inspection of the substation , reducing the inspection workload of the inspectors and improving the intelligent inspection level of the substation, thereby effectively ensuring the safe and stable operation of the substation.

Description of drawings

FIG. 1 is a schematic diagram of a topology structure of a system according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of an interaction principle of a system according to an embodiment of the present invention.

FIG. 3 is a schematic diagram of functional modules of an intelligent operation and maintenance master station in a system according to an embodiment of the present invention.

FIG. 4 is a schematic diagram of a basic flow of a method according to an embodiment of the present invention.

Detailed ways

The image recognition-based substation intelligent patrol system and its application method of the present invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

As shown in FIG. 1 and FIG. 2 , the intelligent inspection system for substations based on image recognition in this embodiment includes an online monitoring sub-station 1 , an online monitoring main station 2 , an intelligent operation and maintenance main station 3 , an image recognition server 4 and a mobile operation and maintenance terminal 5 , the online monitoring sub-station 1, the online monitoring master station 2, and the intelligent operation and maintenance master station 3 are connected in sequence through the network, and the intelligent operation and maintenance master station 3 and the image recognition server 4 are connected through the network, and the intelligent operation and maintenance master station 3, The mobile operation and maintenance terminals 5 are connected through the network, and the input end of the online monitoring sub-station 1 is connected with the monitored secondary equipment to send the state monitoring data of the secondary equipment composed of the collected secondary equipment status and alarm information to the online monitoring system. The monitoring master station 2, the online monitoring master station 2 aggregates the status monitoring data of the secondary equipment of each online monitoring sub-station 1 and pushes it to the intelligent operation and maintenance master station 3, which is used for the mobile operation and maintenance terminal 5 The collected images call the image recognition server 4 to perform image recognition and combine with the state monitoring data of the secondary equipment to generate inspection results and output them to the mobile operation and maintenance terminal 5 .

As shown in FIG. 1 , this embodiment also includes three information security areas separated by physical space: information security area I, information security area II, and information security area III. The online monitoring sub-station 1 and its monitored secondary equipment are located in In the information security area I, the online monitoring master station 2 is located in the information security area II, and the intelligent operation and maintenance master station 3, the image recognition server 4 and the mobile operation and maintenance terminal 5 are located in the information security area III.

In this embodiment, the network between the online monitoring sub-station 1, the online monitoring main station 2, and the intelligent operation and maintenance main station 3 adopts the MMS communication protocol for communication connection, and the input terminal of the online monitoring sub-station 1 is connected to the monitored secondary equipment. The MMS communication protocol is used for communication connection, which is compatible with all kinds of monitored secondary equipment in existing substations.

In this embodiment, the intelligent operation and maintenance master station 3 and the mobile operation and maintenance terminal 5 are connected through a power wireless private network, and the mobile operation and maintenance terminal 5 uses the power wireless private network to upload the collected data to the intelligent operation and maintenance master station 3, The intelligent operation and maintenance main station 3 uses the power wireless private network to send the inspection results to the mobile operation and maintenance terminal 5 .

In this embodiment, a private protocol is used for communication between the intelligent operation and maintenance master station 3 and the image recognition server 4, and the image recognition server 4 can be protected by the above method.

In this embodiment, the online monitoring sub-station 1 mainly implements communication with the secondary equipment in the information security area I, collects the status of the secondary equipment online in real time, and sends the collected secondary equipment status and alarm information to the online equipment in the information security area II. monitoring master station;

In this embodiment, the online monitoring master station 2 mainly realizes the state monitoring data of the online monitoring substations in the integrated information security zone I, and pushes the online state monitoring data to the intelligent operation and maintenance master station in the information security zone III;

Referring to FIG. 3 , in this embodiment, the intelligent operation and maintenance master station 3 should include the following functional modules:

The secondary equipment state library establishment module realizes the establishment of the secondary equipment state of the inspection, which is the basis for the automatic determination of the inspection result;

Recognition secondary equipment object and recognition text matching module, realize the rapid matching of secondary equipment recognition and recognition text, and establish a complete secondary equipment recognition object;

The matching module between the image recognition object and the actual secondary equipment realizes the matching relationship between the secondary equipment recognition object and the object to be inspected;

The inspection result analysis module realizes the automatic determination and output of the inspection results of the secondary equipment according to the established secondary equipment state library and the actual secondary equipment inspection state;

The communication module realizes the communication service between the intelligent operation and maintenance master station, the online monitoring master station, the image recognition server, and the mobile operation and maintenance terminal.

Among them, the secondary equipment state library establishment module, the recognition secondary equipment object and the recognition text matching module, the graphic recognition object and the actual secondary equipment matching module, and the inspection result analysis module are all program modules, and the secondary equipment state library establishment module is used to establish Secondary equipment state library, the matching module for identifying secondary equipment objects and identifying characters is used to perform step 4.2), and the matching module between graphic recognition objects and actual secondary equipment is used to execute step 4.3), and the matching method for identifying secondary equipment objects and identifying characters : Match the rectangular frame of the recognized secondary device object with the rectangular frame of the recognized text, traverse the rectangular frame of the recognized text around it according to the coordinates of the rectangular frame of the recognized secondary device object, and determine the text rectangular frame closest to the rectangular frame of the recognized secondary device object. This recognition character is a character which is a secondary device object for recognition. Image recognition object and actual secondary equipment matching method: use the recognition text as the unique identification of the secondary equipment object, and perform traversal matching with the secondary equipment object to be inspected. The inspection result analysis method of the inspection result analysis module is as follows: according to the running state of each secondary equipment of the substation pushed by the online monitoring master station, automatically select the current applicable secondary equipment state library of the substation, and based on the secondary equipment state library and the actual collected secondary equipment state library. Check the status information of the secondary equipment images. The secondary equipment inspection results with consistent status information are normal, and the secondary equipment inspection results with inconsistent status information are abnormal. The communication module is a hardware module for realizing data communication with the online monitoring master station 2 , the image recognition server 4 , and the mobile operation and maintenance terminal 5 respectively, and specifically includes a plurality of communication interface modules. The intelligent operation and maintenance main station 3 pushes the automatic inspection results to the mobile operation and maintenance terminal 5 through the communication interface module, and the on-site personnel conduct the next secondary equipment inspection or the current secondary equipment inspection results according to the inspection results of the intelligent operation and maintenance main station 3. Manual confirmation .

In this embodiment, the image recognition server 4 has a built-in graphic recognition engine, and the graphic recognition engine has the function of classifying secondary equipment images and their corresponding secondary equipment states. Identify object state. The pattern recognition engine can adopt the existing machine learning classification model as required, and improve the pattern recognition accuracy of the machine learning classification model through pre-training.

In this embodiment, the image recognition server 4 should include the following functional modules:

Secondary equipment identification module, the secondary equipment identification module is integrated with the aforementioned graphic identification engine, which is used to realize the identification of the secondary equipment of the substation and the identification of the status of the secondary equipment;

The text recognition module realizes the recognition of the text associated with the secondary equipment of the substation;

The communication module realizes the communication between the image recognition server and the intelligent operation and maintenance master station, and sends the image recognition result to the intelligent operation and maintenance master station for inspection service processing.

In this embodiment, the mobile operation and maintenance terminal 5 should include the following functional modules:

The image acquisition module realizes the image acquisition of the actual operating state of the secondary equipment of the substation;

The communication module realizes the communication between the mobile operation and maintenance terminal and the intelligent operation and maintenance master station, and sends the actual operation status image of the secondary equipment to the intelligent operation and maintenance master station for intelligent inspection service processing;

The inspection result output module realizes the automatic inspection result display of the mobile operation and maintenance terminal, and displays the business processing results of the intelligent operation and maintenance master station in the mobile operation and maintenance terminal for the operation and maintenance personnel to view.

The online monitoring sub-station and the online monitoring main station are part of the intelligent operation and inspection system of the substation, and real-time online collection of the status of the secondary equipment on site is used as the basis for the operation and maintenance inspection operation.

As shown in FIG. 4 , this embodiment also provides an application method of the aforementioned image recognition-based intelligent substation patrol system. The implementation steps include:

1) Select the target secondary equipment, the intelligent operation and maintenance master station 3 collects the status monitoring data of the target secondary equipment through the online monitoring sub-station 1 and the online monitoring master station 2, and automatically selects the corresponding secondary equipment according to the current operating status of the secondary equipment. The equipment status library, the secondary equipment status library includes the current status of the target secondary equipment; the mobile operation and maintenance terminal 5 uses its own camera to collect on-site inspection images for the target secondary equipment, and upload the inspection images to the intelligent operation and maintenance terminal through the network. Dimension master station 3;

2) The intelligent operation and maintenance master station 3 sends the inspection image to the image recognition server 4;

3) The image recognition server 4 performs image recognition on the patrol image to obtain the image recognition object status of the recognition secondary equipment, and returns the image recognition result to the intelligent operation and maintenance master station 3;

4) The intelligent operation and maintenance master station 3 determines the image recognition object status of the target secondary equipment according to the image recognition object status of the recognition secondary equipment obtained by image recognition, and compares the image recognition object status of the target secondary equipment and its corresponding secondary equipment. The equipment state library checks the state information. If the state of the image recognition object is consistent with the state of the secondary equipment state library, the automatic inspection result is judged to be normal, otherwise it is judged to be abnormal;

5) The intelligent operation and maintenance main station 3 pushes the automatic inspection results to the mobile operation and maintenance terminal 5, and the on-site personnel conduct the next secondary equipment inspection or the current secondary equipment inspection results manually according to the automatic inspection results.

Referring to FIG. 4 , step 1) of this embodiment also includes the step of establishing a secondary device state library, and establishing a secondary device state library is specifically one of the following methods ① and ② or a combination of the two. , Method ①: for each state of the target secondary equipment, obtain the secondary equipment state library of the target secondary equipment in this state by manually establishing the secondary equipment and confirming the status of the secondary equipment; method ②: for the target two For each state of the secondary equipment, the on-site personnel collect the images of the secondary equipment on-site, perform image recognition, and automatically enter the secondary equipment and secondary equipment status to obtain the secondary equipment state library of the target secondary equipment in this state. One or a combination of the two methods of mode ① and mode ② can edit and save the secondary equipment state library on the intelligent operation and maintenance master station 3 as the basis for the analysis of the inspection results.

As an optional implementation manner, the secondary equipment types entered in the secondary equipment state library in this embodiment include: air switch, pressure plate, device indicator light, and the like.

In step 1), the online monitoring sub-station 1 is connected to the online monitoring master station 2, and the online monitoring master station 2 is connected to the intelligent operation and maintenance master station 3 to ensure that the intelligent operation and maintenance master station 3 collects the operation status of the secondary equipment in real time, The intelligent operation and maintenance master station 3 automatically selects the secondary equipment state library according to the current secondary equipment operating state.

In this embodiment, the step of determining the image recognition object state of the target secondary device according to the image recognition object state of the recognition secondary device obtained by image recognition in step 4) includes:

4.1) Obtain the object rectangle for recognizing secondary equipment and the rectangle for recognizing text in the image recognition result;

4.2) Match the rectangular frame of the object for identifying the secondary device and the rectangular frame of the text to obtain the object text for identifying the secondary device;

4.3) Match each recognition secondary device and its object text with the target secondary device to obtain the image recognition result composed of the object rectangular frame and the object text of the target secondary device.

In this embodiment, the detailed steps of step 4.2) include: traversing the rectangular frame of the recognized text around it according to the coordinates of the object rectangular frame for recognizing the secondary device, and determining the rectangular frame that is closest to the recognized text to the object rectangular frame for recognizing the secondary device as this The recognition character is the target character for recognizing the recognition secondary device.

In this embodiment, the detailed steps of step 4.3) include: matching each recognition secondary device and its object text with the target secondary device, so as to obtain an image recognition result composed of the object rectangular frame and the object text of the target secondary device.

To sum up, the application method of the substation intelligent inspection system based on image recognition in this embodiment establishes a secondary equipment state database on the intelligent operation and maintenance main station 3. Station 2 obtains the running status of the secondary equipment online, and automatically selects the current secondary equipment status library for inspection according to the online running status of the secondary equipment. The image is uploaded to the intelligent operation and maintenance main station 3. The intelligent operation and maintenance main station 3 uploads the secondary equipment image to the image recognition server 4, and the image recognition server 4 returns the image recognition result to the intelligent operation and maintenance main station 3. The intelligent operation and maintenance main station 3 combines the inspection of the secondary equipment state library with the Based on the image recognition results, the inspection results are automatically determined, and the inspection results are pushed to the mobile operation and maintenance terminal 5, so that the inspection efficiency and quality can be effectively improved.

The above are only the preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions under the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those skilled in the art, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims (10)

1. A substation intelligent patrol system based on image recognition, characterized in that it comprises an online monitoring sub-station (1), an online monitoring master station (2), an intelligent operation and maintenance master station (3), an image recognition server (4) and A mobile operation and maintenance terminal (5), the online monitoring sub-station (1), the online monitoring main station (2), and the intelligent operation and maintenance main station (3) are connected in sequence through a network, and the intelligent operation and maintenance main station (3) ) and the image recognition server (4) are connected through the network, the intelligent operation and maintenance master station (3) and the mobile operation and maintenance terminal (5) are connected through the network, and the input terminal of the online monitoring sub-station (1) Connect to the monitored secondary equipment to send the state monitoring data of the secondary equipment composed of the collected secondary equipment status and alarm information to the online monitoring master station (2), the online monitoring master station (2) The state monitoring data of the secondary equipment of the monitoring sub-station (1) is aggregated and pushed to the main intelligent operation and maintenance station (3), where the main intelligent operation and maintenance station (3) is used to collect images according to the mobile operation and maintenance terminal (5). The image recognition server (4) is called to perform image recognition and combined with the state monitoring data of the secondary equipment to generate inspection results and output to the mobile operation and maintenance terminal (5). 2. The intelligent patrol system for substations based on image recognition according to claim 1, characterized in that, it also comprises a total of three physical space-isolated information security areas of information security I area, information security II area, and information security III area. The online monitoring sub-station (1) and its monitored secondary equipment are located in the information security area I, the online monitoring main station (2) is located in the information security area II, the intelligent operation and maintenance main station (3), image recognition The server (4) and the mobile operation and maintenance terminal (5) are located in the information security zone III. 3. The substation intelligent patrol system based on image recognition according to claim 1, characterized in that, between the online monitoring sub-station (1), the online monitoring main station (2), and the intelligent operation and maintenance main station (3) The network of the device adopts the MMS communication protocol for communication connection, and the input end of the online monitoring sub-station (1) and the monitored secondary equipment adopt the MMS communication protocol for communication connection. 4 . The intelligent inspection system for substations based on image recognition according to claim 1 , wherein the intelligent operation and maintenance master station ( 3 ) and the mobile operation and maintenance terminal ( 5 ) are connected through a power wireless private network. 5 . 5 . The intelligent inspection system for substations based on image recognition according to claim 1 , characterized in that a private protocol is used for communication between the intelligent operation and maintenance master station ( 3 ) and the image recognition server ( 4 ). 6 . 6 . The application method of the image recognition-based intelligent patrol system for substations according to any one of claims 1 to 5 , wherein the implementing steps include: 1) Select the target secondary equipment, and the intelligent operation and maintenance master station (3) collects the status monitoring data of the target secondary equipment through the online monitoring sub-station (1) and the online monitoring master station (2), according to the current operating status of the secondary equipment The corresponding secondary equipment state library is automatically selected, and the secondary equipment state library includes the current state of the target secondary equipment; the mobile operation and maintenance terminal (5) uses its own camera to collect on-site inspection images for the target secondary equipment, and Send the inspection image to the intelligent operation and maintenance master station through the network (3); 2) The intelligent operation and maintenance master station (3) sends the inspection image to the image recognition server (4); 3) The image recognition server (4) performs image recognition on the patrol image to obtain the image recognition object status of the recognition secondary equipment, and returns the image recognition result to the intelligent operation and maintenance master station (3); 4) Intelligent operation and maintenance master station (3) Determine the image recognition object status of the target secondary equipment according to the image recognition object status of the recognition secondary equipment obtained by image recognition, and compare the image recognition object status of the target secondary equipment and its corresponding The state information of the secondary equipment is checked. If the state of the image recognition object is consistent with the state of the secondary equipment state library, the automatic inspection result is judged to be normal, otherwise it is judged to be abnormal; 5) The intelligent operation and maintenance master station (3) pushes the automatic inspection results to the mobile operation and maintenance terminal (5), and the on-site personnel conduct the next secondary equipment inspection or the current secondary equipment inspection results manually according to the automatic inspection results. 7 . The application method of the intelligent inspection system for substations based on image recognition according to claim 6 , wherein, before step 1), it further comprises the step of establishing a secondary equipment state library, and establishing the secondary equipment state library is specifically as follows: 8 . One or a combination of the two methods mentioned above: Method ① and Method ②. Method ①: For each state of the target secondary equipment, manually establish the secondary equipment and confirm the status of the secondary equipment to obtain the target secondary equipment. The secondary equipment state library of the equipment in this state; method ②: For each state of the target secondary equipment, the on-site secondary equipment images are collected by on-site personnel, and image recognition is performed. The entry of , obtains the secondary equipment state library of the target secondary equipment in this state. 8 . The application method of the intelligent substation patrol system based on image recognition according to claim 6 , wherein in step 4), the image of the target secondary equipment is determined according to the image recognition object state of the secondary equipment recognized by the image recognition. The steps to identify the state of an object include: 4.1) Obtain the object rectangle for recognizing secondary equipment and the rectangle for recognizing text in the image recognition result; 4.2) Match the rectangular frame of the object for identifying the secondary device and the rectangular frame of the text to obtain the object text for identifying the secondary device; 4.3) Match each recognition secondary device and its object text with the target secondary device to obtain the image recognition result composed of the object rectangular frame and the object text of the target secondary device. 9 . The application method of the intelligent inspection system for substations based on image recognition according to claim 8 , wherein the detailed steps of step 4.2) include: traversing the rectangle of the recognized text around it according to the coordinates of the rectangular frame of the object for recognizing the secondary equipment. 10 . frame, the rectangular frame of the recognized character closest to the target rectangular frame for recognizing the secondary device determines that the recognized character is the target character for recognizing the secondary device. 10 . The application method of the image recognition-based intelligent substation patrol system according to claim 8 , wherein the detailed steps of step 4.3) include: matching each identifying secondary device and its object text with the target secondary device, 11 . Thereby, the image recognition result composed of the object rectangular frame and the object text of the target secondary device is obtained.

Images