Remote intelligent inspection solution for substations

1、 Scheme Introduction

With the rapid development of the power grid in recent years, the application level of visualization devices has continued to deepen, and the demand for equipment visualization in grassroots units has been constantly increasing. The contradiction between production personnel and inspection workload is becoming increasingly prominent, and the integration of operation and maintenance, as well as the lean transformation of substations, have raised higher requirements for substation operation and inspection work while expanding their business scope. In this situation, the traditional "manual inspection and manual recording" inspection operation method is difficult to adapt to the development requirements of lean and intensive power systems. The distribution range of substations is relatively wide, requiring a large amount of data for daily inspections, checks, and records. There are few on-site personnel, resulting in a large workload. In addition, with the continuous improvement of equipment operation and maintenance lean requirements by State Grid, there is an urgent need to enhance the intelligence level of daily inspections.

The remote intelligent inspection and control system for substations applies big data analysis, artificial intelligence, image recognition, and automatic navigation technology to multi-dimensional high-definition remote intelligent inspections of protection screens, production equipment, and production environments through means such as "deepening the application of robots, upgrading high-definition video functions, intelligent linkage of auxiliary control systems, and unmanned aerial vehicle inspections". It also supports remote operation and video confirmation of equipment, linkage analysis of scheduling actions, abnormalities, faults, and accident video recordings, visualization of remote equipment maintenance diagnosis, and real-time positioning and safety management of operators, achieving the replacement of manual labor by machines, improving the lean and intelligent level of intelligent operation and inspection, enhancing the work efficiency and quality of operation and maintenance personnel, and strengthening equipment status control.

2、 Program features

1. Remote inspection visualization: Through high-definition network cameras and inspection robots, equipment can be monitored from all angles and multiple perspectives. Through remote automatic inspection and remote manual inspection, comprehensive inspection, routine inspection, light off inspection, special inspection and other work can be completed, and inspection results can be automatically recorded and inspection reports generated.

2. Automatic recognition of inspection defects: The system adopts image comparison, artificial intelligence and other image and video analysis technologies to achieve equipment status recognition and equipment appearance defect recognition, establish an equipment appearance defect library, and improve defect recognition accuracy through continuous automatic collection and learning of samples, meeting the needs of substation equipment inspection.

3. Abnormal intelligent warning: The system obtains data such as main transformer oil temperature, oil level, SF6 pressure gauge, and temperature of equipment heating parts, forms historical data curves, automatically analyzes the trend and correlation of data changes, and realizes intelligent judgment of equipment operation status. When the development trend of these data may affect the normal operation of power equipment, the system provides early warning to remind operation and maintenance personnel to pay close attention or conduct on-site inspections.

4. Intelligent linkage between main and auxiliary equipment: The system forms an alarm linkage with the main and auxiliary equipment monitoring system. After receiving signals such as remote control preset of the main equipment, displacement of the main and auxiliary equipment, and exceeding the limit or alarm of the main and auxiliary equipment monitoring system, according to the corresponding relationship between the configured linkage signal and the inspection point, it automatically generates video inspection tasks to inspect the points that need to be reviewed, which facilitates comprehensive analysis and judgment of the site by operation and maintenance personnel.

3、 Scheme system architecture

The remote intelligent inspection and control system for substations is deployed at the substation end, mainly composed of inspection hosts (including intelligent application units and intelligent analysis units), station end intelligent inspection workstations, station control layer switches, video hosts, high-definition videos, etc. The inspection host issues commands such as control and inspection tasks, and the video host controls the camera to carry out joint indoor and outdoor equipment inspection operations, uploading inspection data, collected files, etc. to the inspection host; The inspection host intelligently analyzes the collected data to generate inspection results and inspection reports. It has functions such as real-time monitoring and intelligent linkage with the main and auxiliary monitoring systems. The inspection host has four network ports and independent network segments, and its information security meets the requirements of the "National Development and Reform Commission Order No. 14 on Security Protection of Power Monitoring Systems".

The system architecture is shown in the following figure:

4、 Scheme system functions

1. Automatic data collection

(1) Patrol data collection

Visible light photo data collection: Collect readings from SF6 pressure gauge, switch action count counter, lightning arrester leakage current gauge, oil temperature gauge, hydraulic gauge, on load voltage regulating gear gauge, oil level gauge, and other gauges; Collect position and status indicators of primary equipment such as circuit breakers and isolation switches, as well as secondary equipment such as switching handles, pressure plates, indicator lights, and air switches; The appearance and other conditions of the collection equipment and facilities.

Infrared spectrum data collection: including the number of infrared spectra of key parts such as equipment body, joints, sleeves, and leads

Audio data collection: including sound data of equipment such as transformers and reactors.

(2) Robot state data collection

(3) Environmental data collection

Collect micro meteorological data such as atmospheric temperature, humidity, wind speed, and rain and snow conditions through micro meteorological equipment.

2. Intelligent image analysis

Defect recognition: Using deep learning intelligent analysis technology, automatically identify equipment defects such as dial blurring, meter damage, insulator breakage, silicone discoloration, oil seal damage, hanging suspended solids, bird nests, component surface oil stains, ground oil stains, metal corrosion, etc,Analyze whether there are abnormalities in the image, and label and save the suspected abnormal images with rectangular boxes on the images.

State recognition: The system can intelligently analyze infrared spectra, visible light photos, and other data recorded during the joint inspection process. By analyzing the infrared spectrum, obtain the highest temperature information of the target device in the image, and automatically determine whether the device has overheating defects; Through image recognition technology, automatically determine the operating status of equipment, including meter readings and switch device position indication status.

3. Real time monitoring display

Patrol monitoring realizes the display of robot status information, patrol map display, and video monitoring of robots and video monitoring systems; Robot control realizes pan tilt control operation, setting and calling of pan tilt preset positions, robot self check, remote reset, one click return and other control functions; The video surveillance system controls the pan tilt control, visible light image control, infrared image control, audio control and other control functions of the video surveillance system.

4. Inspection task management

Implement the functions of setting, issuing, and displaying inspection tasks. The content of inspection tasks includes information on inspection points and inspection cycles in different time dimensions such as month, week, day, and hour. Inspection types include comprehensive inspection, routine inspection, light off inspection, special inspection, special inspection, custom inspection, etc6Class. Task execution supports three methods: immediate execution, regular execution, and periodic execution.

5. Automatic intelligent inspection

The system operates autonomously, with robots and video cameras executing inspection tasks according to pre designed inspection routes, conducting intelligent inspections of substations and promptly detecting equipment problems, thermal faults, appearance defects, and other abnormal phenomena, resulting in high operational efficiency and quality.

6. Confirmation of Inspection Results

Inspection result browsing: Implement the confirmation and archiving of inspection data, automatically save the results after the inspection is completed, and display them in the form of a device tree according to the order of inspection point settings. Each inspection point contains all the collected information and thresholds of the current inspection task. After all points are confirmed, record the reviewer and audit time, and have the function of identifying abnormal or unidentifiable inspection points for summary and query.

Equipment alarm confirmation: provides equipment alarm analysis, alarm query confirmation, main wiring display, interval display functions, including equipment name, component name, defect category, alarm level, defect or abnormal image (with specific defect or abnormal location marked) and other information; The real-time monitoring screen link of the alarm information can be quickly accessed, and the real-time monitoring screen of the alarm device can be manually viewed to verify the authenticity of the alarm information and input feedback opinions.

Inspection report generation: The system has query, browsing, and output functions. Automatically generate inspection reports after completing the inspection task review; Capable of querying historical inspection reports by combining conditions such as equipment area, interval name, equipment type, detection type, and inspection time period; Task query has the functions of querying, resetting, exporting, and viewing reports.

7. Analysis of Inspection Results

Historical data analysis: The system integrates historical inspection data of various devices to form a historical status data curve of the equipment, which more intuitively reflects the trend of equipment status changes. By analyzing the curve fluctuations and trends between different equipment data in different time periods, it macroscopically controls common faults and their causes, thereby achieving the goal of early detection, analysis, and solution of faults.

Trend analysis: The system has intelligent analysis function, which can effectively integrate and correlate inspection data, automatically analyze its changing trend and correlation relationship, and automatically alert when the data reaches the alarm threshold. When the development trend of these data may affect the normal operation of power equipment, the system will give early warning, extract valuable information from the trend and characteristics of historical data, remind operation and inspection personnel to pay close attention to or conduct on-site inspections, and prevent hidden dangers in equipment operation.

8. Intelligent linkage for patrol inspection

The system has the linkage function with the main device's remote control preset signal, displacement signal, over limit signal, and alarm signal; The main and auxiliary monitoring systems send linkage signals to the inspection host. After receiving the linkage signal, the inspection host automatically generates inspection tasks based on the configured linkage signal and the corresponding relationship between the inspection points. Robots or videos are used to inspect the points that need to be reviewed. After the review point inspection is completed, the review results can be viewed on the inspection host.