CN-121999594-A - Intelligent abnormal behavior identification and safety prevention and control system under charging energy scene

CN121999594ACN 121999594 ACN121999594 ACN 121999594ACN-121999594-A

Abstract

The invention discloses an abnormal behavior intelligent identification and safety prevention and control system in a charging energy scene, which comprises a multi-mode data acquisition module, an abnormal behavior intelligent identification module, a risk early warning and decision module and a risk early warning and decision module, wherein the multi-mode data acquisition module is used for acquiring electric data, environment data, video data and space association data in the charging scene in real time, the abnormal behavior intelligent identification module is used for carrying out fusion analysis on the data acquired by the multi-mode data acquisition module, identifying electric abnormality, equipment abnormality, personnel behavior abnormality and environmental risk based on a deep learning model, and the risk early warning and decision module is used for carrying out risk assessment on the identified abnormal behavior.

Inventors

LI CONG
MA ZHEXUAN

Assignees

北京电力经济技术研究院有限公司
中环低碳节能技术(北京)有限公司

Dates

Publication Date: 20260508
Application Date: 20260316

Claims (10)

1. Abnormal behavior intelligent identification and safety prevention and control system under charging energy scene, its characterized in that includes: the multi-mode data acquisition module is used for acquiring electrical data, environment data, video data and space associated data in a charging scene in real time; The abnormal behavior intelligent recognition module is used for carrying out fusion analysis on the data acquired by the multi-mode data acquisition module and recognizing electrical abnormality, equipment abnormality, personnel behavior abnormality and environmental risk based on a deep learning model; The risk early warning and decision module is used for carrying out risk assessment on the identified abnormal behaviors, generating grading early warning information and a prevention and control decision instruction, introducing a risk assessment unit, constructing a quantization model of a comprehensive risk value based on the abnormal type, duration and abnormal strength, generating grading early warning from one level to four levels by calculating the comprehensive risk value and mapping the comprehensive risk value to a preset risk level threshold interval, and outputting a matched prevention and control decision instruction; the linkage prevention and control execution module is used for receiving the prevention and control decision instruction, executing corresponding electric control, fire extinguishing and alarm notification operations and continuously collecting feedback data after execution; The risk early warning and decision module is further configured to start an effect verification mechanism after a prevention and control instruction is sent, acquire feedback data returned by the linkage prevention and control execution module in real time, judge that the risk is not eliminated if the abnormal strength is detected not to be reduced or the duration time is continuously increased, recalculate the comprehensive risk value, dynamically improve the risk level according to the updated comprehensive risk value, trigger higher-level early warning and more severe prevention and control measures, and form closed-loop dynamic optimization control.
2. The intelligent recognition and safety prevention and control system for abnormal behavior in a charging energy scene according to claim 1, wherein the multi-modal data acquisition module comprises: The electric monitoring unit is configured to monitor the voltage, current, power, electric quantity and arc signals of the charging equipment, and the sampling rate is not lower than 10kHz; the environment sensing unit is configured to acquire temperature, smoke concentration, combustible gas concentration and humidity data of the charging station; The video behavior monitoring unit is configured to identify personnel operation behaviors, vehicle states and visible light characteristics of open fire and smoke through video images; the space association sensing unit is configured to acquire operation association data among a plurality of charging devices in the same charging field and construct a device relationship map.
3. The system for intelligent recognition and safety prevention and control of abnormal behavior in a charging energy scene according to claim 2, wherein the intelligent recognition module for abnormal behavior comprises: The data fusion unit is configured to perform space-time alignment and feature extraction on the electrical data, the environmental data, the video data and the space-related data to generate fusion feature vectors; And the abnormal recognition unit is configured to input the fusion feature vector into a deep learning model and recognize abnormal behaviors, wherein the abnormal behaviors comprise overload, short circuit, arc fault, battery thermal runaway, personnel falling and illegal operation.
4. The system for intelligent recognition and safety prevention and control of abnormal behavior in a charged energy scene according to claim 3, wherein the deep learning model adopts a convolutional neural network in combination with a concept enhancement technique, predefines key concepts related to the abnormal behavior of the charging field and performs feature enhancement.
5. The system for intelligent recognition and safety prevention and control of abnormal behavior in a charged energy scene according to claim 3, wherein the intelligent recognition module of abnormal behavior is further configured to migrate knowledge from a common abnormal sample by using a small sample learning technique through a meta learning framework to improve recognition accuracy of rare abnormal behavior.
6. The intelligent recognition and safety prevention and control system for abnormal behavior in a charging energy scene according to claim 1, wherein the risk early warning and decision module comprises: The risk assessment unit is configured to calculate a risk level according to the type, duration and intensity of the abnormal behavior; The early warning generation unit is configured to generate early warning information of different levels according to the risk level and display the information through a visual interface, wherein the visual interface adopts a charging station three-dimensional model constructed based on a BIM or GIM technology to calibrate abnormal positions and states; and the decision instruction unit is configured to generate a control instruction corresponding to the risk level according to a preset control strategy.
7. The intelligent recognition and safety prevention and control system for abnormal behavior in a charging energy scene according to claim 2, further comprising a system management platform for providing data storage, model training, man-machine interaction and system configuration functions and coordinating the modules to work cooperatively.
8. The system for intelligently identifying and safely preventing and controlling abnormal behaviors in a charging energy scene according to claim 7, wherein the system management platform is integrated with a digital twin module, and the digital twin module is configured to: Constructing a virtual mapping model of the charging station equipment; based on the real-time data and the historical data, the execution effect of the control instruction is simulated and optimized.
9. The intelligent recognition and safety prevention and control system for abnormal behavior in a charging energy scene according to claim 1, wherein the linkage prevention and control execution module comprises: the electric prevention and control unit is configured to execute circuit on-off control, charging power adjustment or active arc breaking operation; A fire extinguishing unit configured to activate an extinguishing device for a lithium battery fire when a risk of the fire is identified; and the alarm notification unit is configured to send alarm information to the mobile terminal of the operation and maintenance personnel and the monitoring center and start the on-site audible and visual alarm equipment.
10. The system for intelligent recognition and security prevention and control of abnormal behavior in a charged energy scenario of claim 8, wherein the system management platform is further configured to: Receiving feedback information of operation and maintenance personnel aiming at an abnormal processing result; Based on the feedback information, a machine learning algorithm is utilized to carry out self-adaptive optimization updating on the recognition model in the abnormal behavior intelligent recognition module or the decision rule in the risk early warning and decision module.

Description

Intelligent abnormal behavior identification and safety prevention and control system under charging energy scene Technical Field The invention relates to the technical field of new energy technology, safety prevention and control of intelligent charging facilities and artificial intelligence intersection, in particular to an abnormal behavior intelligent identification and safety prevention and control system under a charging energy scene. Background Currently, most charging facilities still rely on a traditional threshold alarming mechanism and an independent video monitoring system to conduct safety supervision, and particularly under complex working conditions, single sensor data is easy to interfere, and the false alarm rate is high. For example, chinese patent No. 120840446A discloses a safety monitoring system of an automobile charging pile based on multi-mode sensor fusion, and in particular discloses a safety monitoring system of an automobile charging pile based on multi-mode sensor fusion, which comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor realizes the following steps when executing the computer program, according to the data difference of each monitoring index of any automobile charging pile to be detected and an automobile charging pile with normal working condition, the electric abnormality degree of any automobile charging pile to be detected is obtained. The patent uses a rule engine to independently judge the out-of-limit conditions of various parameters by collecting current, voltage, temperature and image data and triggers corresponding alarm. Because the existing system is mainly controlled in an open loop manner, a risk closed loop feedback mechanism is not established, namely whether the risk is eliminated or not can not be confirmed after an alarm is sent out, if a disposal measure (such as power failure) is invalid, the system can not automatically upgrade a response, so that the risk is out of control, and risk judgment can not be dynamically adjusted according to state data after prevention and control execution. Disclosure of Invention Therefore, the invention provides an intelligent abnormal behavior identification and safety prevention and control system under a charging energy scene so as to solve the problems in the prior art. In order to achieve the above object, the present invention provides the following technical solutions: abnormal behavior intelligent identification and safety prevention and control system under charging energy scene includes: the multi-mode data acquisition module is used for acquiring electrical data, environment data, video data and space associated data in a charging scene in real time; The abnormal behavior intelligent recognition module is used for carrying out fusion analysis on the data acquired by the multi-mode data acquisition module and recognizing electrical abnormality, equipment abnormality, personnel behavior abnormality and environmental risk based on a deep learning model; The risk early warning and decision module is used for carrying out risk assessment on the identified abnormal behaviors, generating grading early warning information and prevention and control decision instructions, introducing a risk assessment unit, constructing a quantization model of a comprehensive risk value based on the abnormal type, duration and abnormal strength, generating grading early warning from one level to four levels by calculating the comprehensive risk value and mapping the comprehensive risk value to a preset risk level threshold interval, and outputting the matched prevention and control decision instructions: the linkage prevention and control execution module is used for receiving the prevention and control decision instruction, executing corresponding electric control, fire extinguishing and alarm notification operations and continuously collecting feedback data after execution; The risk early warning and decision module is further configured to start an effect verification mechanism after a prevention and control instruction is sent, acquire feedback data returned by the linkage prevention and control execution module in real time, judge that the risk is not eliminated if the abnormal strength is detected not to be reduced or the duration time is continuously increased, recalculate the comprehensive risk value, dynamically improve the risk level according to the updated comprehensive risk value, trigger higher-level early warning and more severe prevention and control measures, and form closed-loop dynamic optimization control. Preferably, the multi-mode data acquisition module includes: The electric monitoring unit is configured to monitor the voltage, current, power, electric quantity and arc signals of the charging equipment, and the sampling rate is not lower than 10kHz; the environment sensing unit is configured to acquire temperature, smoke concentratio