CN-121982599-A - Security operation behavior recognition system and method for heating station based on artificial intelligent image recognition

CN121982599ACN 121982599 ACN121982599 ACN 121982599ACN-121982599-A

Abstract

The application discloses a thermodynamic station safe operation behavior recognition system and method based on artificial intelligent image recognition, and relates to the technical field of industrial safety monitoring. The system comprises an image acquisition unit, a processing unit and an alarm unit, wherein the image acquisition unit comprises various cameras, a key area of a heating power station is covered, the processing unit runs an AI image recognition algorithm, inspection personnel detection, on-duty off-duty recognition, protective equipment inspection and environmental risk monitoring are realized, and the alarm unit automatically triggers an alarm based on a recognition result. The method comprises real-time video acquisition, AI behavior analysis, risk identification and alarm generation. According to the intelligent park platform, unsafe behaviors are automatically identified through an AI algorithm, the instantaneity and the reliability of safety management of a heating power station are remarkably improved, and the system can be expanded and integrated to the intelligent park platform to support multidimensional data display and remote management. The application is suitable for dangerous places such as heating stations and the like, and can effectively prevent safety accidents.

Inventors

DUAN XIXI
HUANG DA
BAI XIAOLEI
WANG ZHIYONG
LIU YUAN
SHU YU
Yang Hailiu
WANG YOUYOU

Assignees

中国空间技术研究院

Dates

Publication Date: 20260505
Application Date: 20251202

Claims (10)

1. A thermodynamic station safe operation behavior recognition system based on artificial intelligence image recognition, comprising: An image acquisition unit configured to acquire video image data within the thermal station; the processing unit is in communication connection with the image acquisition unit and is configured to run an artificial intelligent image recognition algorithm so as to recognize preset safety behaviors and unsafe behaviors; An alarm unit configured to generate and output alarm information in response to the recognition of the unsafe behavior; the artificial intelligent image recognition algorithm comprises a plurality of behavior recognition models and is used for analyzing the on-duty, inspection, operation behaviors and environmental states of operators of the heating power station in real time.
2. The system of claim 1, wherein the image acquisition unit comprises multiple types of cameras selected from at least one of an explosion proof gun camera, a gun camera, and a mini camera and configured to cover critical areas of a thermal station, including boiler channels, gas rooms, heat exchange stations, and laboratory rooms.
3. The system of claim 1, wherein the processing unit comprises a video intelligent analysis platform configured to: managing a plurality of artificial intelligence image recognition algorithms and matching the algorithms to corresponding cameras; Providing a batch operation function of the algorithm task, including starting, stopping, deleting and editing; the intelligent park digital twin platform is integrated to the intelligent park digital twin platform and used for displaying alarm states in a three-dimensional map.
4. The system of claim 1, wherein the artificial intelligence image recognition algorithm comprises a patrol personnel detection algorithm configured to monitor whether patrol personnel arrive at a specified location within a specified time within a preset patrol area, and to trigger an alarm when no personnel are detected; Preferably, the patrol personnel detection algorithm is associated with a patrol identification area plan, the planned area being defined by virtual boundaries and configured based on the area division indicated by the patrol identification.
5. The system of claim 1, wherein the artificial intelligence image recognition algorithm comprises an on-Shift detection algorithm configured to monitor the number of people in the thermal station by a people counting function of the entrance camera, and to trigger an alarm when the number of people is less than a preset threshold.
6. The system of claim 1, wherein the artificial intelligence image recognition algorithm comprises a designated area operation recognition algorithm configured to recognize whether a person is wearing protective equipment correctly during a dosing, water quality testing operation, the protective equipment comprising at least one of acid and alkali resistant gloves, protective gloves, goggles, protective clothing, and a mask; Preferably, the artificial intelligence image recognition algorithm further comprises an environment detection algorithm selected from at least one of a flame recognition algorithm, a deposit recognition algorithm, and a smoke recognition algorithm configured to detect open fires, abnormal deposits, and offending smoke behavior within the thermal station in real time.
7. The system of claim 1, wherein the processing unit comprises a server configured with a GPU acceleration module for improving algorithm reasoning efficiency; Preferably, the reasoning time of the artificial intelligent image recognition algorithm is not more than 0.2 seconds, and the accuracy is not less than 90%.
8. The system of claim 1, further comprising a historical alert management module configured to record, query and derive alert information, support multi-dimensional analysis of time ranges, alert types and alert states, and enable data sharing in conjunction with the intelligent campus platform.
9. The system of claim 1, wherein the video data of the image acquisition unit is stored in a hard disk video recorder configured to meet a storage time of greater than 90 days, and wherein the power and data transmission of the Poe is implemented using a network switch.
10. A method for identifying safety operation behaviors of a heating station based on artificial intelligent image identification, which is realized based on the system of any one of claims 1 to 9, and is characterized by comprising the following steps: Step S1, acquiring a video stream in a heating power station in real time through image acquisition equipment; s2, analyzing the video stream by using an artificial intelligent image recognition algorithm, and recognizing personnel behaviors and environmental risks; And step S3, when unsafe behavior is identified, automatically generating an alarm and notifying relevant personnel.

Description

Security operation behavior recognition system and method for heating station based on artificial intelligent image recognition Technical Field The application relates to a thermodynamic station safe operation behavior recognition system and method based on artificial intelligent image recognition, and belongs to the technical field of thermodynamic station safe management and artificial intelligent image recognition. Background The heating power station is used as a key facility in an energy supply system, belongs to a secondary dangerous point, and is directly related to the continuity of energy supply and the safety of surrounding environment in safe and stable operation. With the high-quality development of the national aerospace industry and related industries, the safety management standard is continuously improved, and the traditional safety supervision mode of the heating power station is gradually exposed to obvious defects. The existing security inspection and maintenance of the heating power station mainly depends on personal experience of operators and traditional video monitoring, and lacks intelligent technical support. In the daily duty scene, the intelligent supervisory personnel cannot be in an on duty state through technical means, the daily inspection quality depends on manual supervision, the influence of human factors is great, the punctuality and the comprehensiveness of inspection are difficult to ensure, the wearing condition of personnel protection equipment is all self-conscious, the personal injury risk is prominent in the operation scene such as dosing, water quality testing and the like, the environmental monitoring depends on manual inspection, the potential safety hazards such as fire disaster, abnormal accumulation and the like are difficult to discover in time, and the problems can cause safety accidents to happen, so that the safety operation of a heating station is seriously influenced. Although the AI image recognition technology has been applied in some industrial scenarios (such as smart power plants and smart sites) to realize preliminary recognition of personnel illegal behaviors and equipment faults, the prior art lacks a customized solution for a specific working scenario of a thermal station (such as water quality test protection, salt tank dosing operation and dedicated inspection route of the thermal station), and cannot fully adapt to the safety management requirement of the thermal station. Therefore, a safe operation behavior recognition system which is specially designed for a thermal station scene, high in intelligent degree and accurate in recognition is needed to solve the defects of the prior art. Disclosure of Invention The invention aims to provide a thermodynamic station safe operation behavior recognition system and method based on AI image recognition, which are used for solving the problems that the traditional supervision mode depends on manpower and has poor real-time performance. The system realizes comprehensive monitoring of personnel behaviors and environmental states of the thermal station through integrating image acquisition, AI algorithm processing and intelligent alarm. According to one aspect of the application, a thermodynamic station safe operation behavior recognition system based on artificial intelligent image recognition is provided, the system comprises an image acquisition unit, a processing unit and an alarm unit, and a history alarm management module can be additionally arranged according to requirements, and the system comprises the following specific steps: Wherein the image acquisition unit is configured to acquire video image data within the thermal station, including multiple types of cameras, selected from at least one of an explosion-proof gun camera, a gun camera, and a small camera. The camera covers key areas of the heating power station according to a preset plan, including boiler channels, gas rooms, heat exchange stations, test rooms, salt tank areas and the like, so that no monitoring blind area is ensured. Preferably, the highest resolution of the explosion-proof gun-type camera can reach 400W, the explosion-proof gun-type camera has a starlight level infrared night vision function, a protection level IP68 is provided, a shell is made of stainless steel 316L material, the explosion-proof level requirement is met, the explosion-proof gun-type camera is installed in dangerous areas such as a boiler room and the like, the highest resolution of the gun-type camera can reach 400W, the starlight level infrared night vision function is installed by adopting a bracket wall hanging and is used for covering a dosing operation area such as a salt tank and the like, the highest resolution of the small-sized camera can reach 400W, and a controllable power supply is arranged by adopting an adherence installation and is used for clearly monitoring the operation process of water quality testing personnel; Preferably, the matched