基于人机协同的铁路客运安检系统研究

Research on Railway Passenger Transport Security Check System Based on Man-computer Collaboration

铁路客运安检是铁路运营安全风险防范工作的重要一环,既有安检仪单机作业,模式落后;值机判图员水平参差不齐,且与开包处置员之间联动方式单一,各岗位作业质量监管考核以及事后回溯困难,缺乏一套满足安检各岗位无缝联动并与智能技术辅助判图相互协同的安检系统,因此,研究设计基于人机协同的铁路客运安检系统。阐述铁路客运安检现状和问题,从总体设计、安检作业流程设计、核心模块功能设计方面进行系统描述;介绍系统功能关键技术,重点论述基于人机协同的安检作业增量学习技术,以提高安检智能识别技术水平和安检作业整体质量与效率;通过南昌站试点应用,证明系统满足安检作业需求,实现了安检各岗位工作质量量化考核,以智能判图辅助监督值机判图员,以人工判图处置结果反馈智能判图模型进行迭代升级,以物防技防促进客运安检提质增效。

Railway passenger transport security check is an important part of railway operation safety risk prevention.The existing security check scanner operates offline in an outdated mode;the on-duty personnel are in different proficiency level of judging the inspection images,and are not coordinated with manual luggage inspectors well,so it is difficult to supervise and assess the operation quality of each post and trace back after events.There is a lack of a security check system that meets the requirements of seamless coordination of each post in security check and the collaboration of intelligent technologyassisted image judgment.Therefore,this paper designs the railway passenger transport security check system based on mancomputer collaboration,expounds the present situation and problems of railway passenger transport security check,and describes the system from the aspects of overall design,security check operation process design and core module function design.This paper introduces the key technologies of system functions,especially elaborates the incremental learning technology based on man-computer collaboration for security check operations to improve the level of intelligent identification technology and the overall quality and efficiency of security check operations.Through the pilot application at Nanchang Railway Station,it is proved that the system meets the needs of security check operation,realizes quantitative assessment of the work quality of each post in security check,assists in supervising the on-duty personnel in judging the inspection images with intelligent image judgment,and feeds back the results of manual image judgment and treatment to the intelligent image judgment model for iterative upgrading,so as to promote the improvement of passenger transport security check quality and efficiency through protection by facilities and technologies.