Research on the application of multi-source sensing technology in terminal safety management

introduction

The multi-source perception intelligent security system is mainly based on the "Petroleum and Petrochemical System Security Counter-Terrorism Prevention Requirements" (GA1551-2019). It targets key security counter-terrorism prevention targets such as oil and gas enterprises and flammable and explosive oil and gas processing stations. According to the overall layout of the plant, it targets key prevention areas such as the perimeter of the terminal processing plant, the entrances and exits of the terminal processing plant, the central control room of the terminal processing plant, and the core device area of ​​oil and gas processing. It combines new technologies such as intelligent video acquisition, intelligent access control, perimeter security monitoring, anti-theft alarm, image recognition, and three-dimensional video fusion to build a multi-source perception "integrated" integrated platform to achieve functional diversification, high-definition equipment, intelligent management, and security visualization. Through the construction of the enterprise production park intelligent security system, the security prevention capability of the enterprise production park will be greatly enhanced, and the management efficiency will be further improved. This paper mainly conducts research from the aspects of multi-source perception intelligent security system architecture, intelligent security technology application, and intelligent security standard construction.

1 Background of Multi-source Perception Security Construction

With the rapid development of technological innovation and information integration such as China Manufacturing, Internet of Things, mobile applications, and artificial intelligence, the strategic position of informatization has become prominent, especially networking, digitization, and intelligence, which will become an inevitable trend of the times. At present, enterprises are in a period of preparation for a new round of scientific and technological revolution and industrial transformation, which puts higher requirements on existing business and management. Therefore, it is becoming more and more important to combine the business scenarios of oil and gas exploration and development, apply information technology to promote and lead innovation and development, and ensure the safe production and safe and stable operation of enterprises.

In recent years, the CPC Central Committee and the State Council have attached great importance to the security and anti-terrorism prevention work in the petroleum and petrochemical industry, and have made a series of major decisions and deployments. The "Requirements for Security and Anti-terrorism Prevention in the Petroleum and Petrochemical System" (GA1551-2019), compiled and issued by the Ministry of Public Security, is an important measure to implement the decisions and deployments of the CPC Central Committee and the State Council. It stipulates that from July 1, 2019, all new construction, reconstruction and expansion projects should include security prevention systems in the overall plan, and be designed, constructed and operated simultaneously. The release and implementation of the "Requirements for Security and Anti-terrorism Prevention in the Petroleum and Petrochemical System" (GA1551-2019) provides key and basic technical support for the security construction of important oil and gas facilities, and has great practical significance for petroleum and petrochemical enterprises to enhance their security prevention capabilities, achieve safe development of the industry, and ensure national energy security and social stability.

2 Main contents of the construction plan

According to the "Petroleum and Petrochemical System Security Counter-Terrorism Prevention Requirements" (GA1551-2019), starting from the three aspects of plant physical prevention, technical prevention, and human prevention, based on the three levels of normal three-level prevention, normal two-level prevention, and normal one-level prevention, as well as abnormal prevention requirements, security counter-terrorism prevention work and management are carried out for key targets and key areas of security counter-terrorism in oil and gas field enterprises. The specific construction framework of the intelligent security solution is shown in Figure 1.

2.1 Intelligent security physical prevention

Anti-collision piles are mainly installed at the entrances of the factory area and production operation area. The anti-collision piles have an automatic lifting function. The lifting of the piles can be controlled in the control room. When authorized vehicles or special vehicles such as fire trucks pass by, the piles can be quickly retracted. Metal fences and other protective measures are installed around the perimeter of the independent factory area, and a security monitoring center and anti-theft doors are installed.

According to the actual situation of the factory and combined with the prevention requirements, a factory security visualization center is built, a security visualization system is deployed, and customized development of video management, alarm, video stitching, linkage, 3D model visualization and other functions.

2.2 Application of Intelligent Security Technology

Cameras are deployed at the perimeter and key areas of the factory to achieve full coverage of video surveillance without blind spots: video surveillance devices are installed at the perimeter entrances and exits, and the video images can clearly show the physical characteristics and license plate information of the people entering and leaving: At the same time, explosion-proof cameras are added to the production area to enhance the ability to collect video information: research is carried out on the application of the factory perimeter intrusion alarm protection system and the anti-UAV active defense system in enterprise safety production. The schematic diagram of intelligent security video surveillance is shown in Figure 2.

(1) Application of video fusion technology.

The existing video surveillance system has the following problems: isolated images, unable to get a global overview; scattered images, unable to determine spatial position; screen constraints, limited observation paths; cross-lens tracking is difficult, and it is overwhelming, etc. Therefore, the video acquisition panoramic fusion technology is used to achieve the integration of video surveillance images in important areas of the factory and improve video tracking efficiency.

(2) Application of 3D visualization technology.

The video surveillance system is combined with the three-dimensional model of the factory area and the camera points to stitch the video of the entire factory area, forming a complete panoramic video visualization system without blind spots. The data collected by the camera and the acquisition device are integrated through the factory security system. The alarm information is reminded in real time through the visualization system, and the voice alarm system is used to give voice alarms to relevant personnel and vehicles. Combined with the video alarm information, the video acquisition program is triggered to record the short video capture of the alarm point and time, which is convenient for playback and retrieval.

Using deep learning and analysis technology on the video platform, we can achieve restricted area detection, crossing detection, human detection, staff feature detection (helmet wearing, etc.), thermal imaging and temperature monitoring, and instrument data acquisition, so as to achieve the purpose of active defense and data acquisition and analysis, provide intelligent technical support for production safety, public security and anti-terrorism, and achieve early warning of accidents and improve the efficiency of command and dispatch. Real-time online extraction of monitoring index requirements, video linkage and alarms, and display and analysis through the system, which improves the information processing efficiency of video monitoring. Actively discover and record on-site safety control indicators, change the original mode of retrieving and analyzing massive video information after problems occur, avoid huge investment in personnel and time, and change from "passive investigation after the event" to "active warning before the event", realize the efficient operation of the on-site video monitoring command system, and improve the on-site safety prevention level of the enterprise's production plant.

(3) Active identification and early warning of safe areas.

Key area monitoring and early warning mainly involves identification, analysis and early warning of personnel entering the factory or safety area, and alarms are issued in combination with electronic fences and sound and light control equipment. Feature recognition technology is used to identify the characteristics of personnel entering the work area, mainly focusing on the wearing of safety helmets and clothing colors. At the same time, video capture is carried out in combination with the video surveillance platform, and abnormal monitoring images are prompted through pop-up windows, changing passive monitoring to active monitoring.

(4) Equipment information image recognition and data acquisition.

Using high-definition visible light cameras and video image recognition technology, the data of the indication and position of meters in the production area, including pressure gauges, temperature gauges, digital displays, liquid level gauges, bearing oil levels, electrical switches and valve status, etc., are read, automatically recorded and generated reports, and alarms are issued after calculation and judgment. At the same time, real-time photos and videos are sent to the background to meet the needs of operators to call, view and automatically archive at any time. It has sVPattenMax image registration technology for intelligent discrimination and analysis of instrument contrast images, which can effectively deal with interference caused by changes in lighting, shadows, occlusions, low contrast and low resolution, changes in viewing angles and scale scaling. It can achieve high-level accurate recognition of key equipment components, and combined with visual recognition technology, it can realize target recognition and locking, intelligent analysis of target readings, and subsequent zooming to obtain high-quality pictures, which improves the recognition accuracy.