摘要
基于电梯设备的广泛安装产生的巨量维保需求,以社会生活密切相关的电梯运行安全为场景、综合验证AI人工智能技术在国产边缘设备实现电梯视频数据采集、AI模型快速开发、AI系统能力中台技术。研究项目包括:5G技术下的电梯运行数据(传感器和视频)传输的经济效益和用户体验、电梯安全AI模型开发(火情、电动车检测)、基于国产芯片的边缘处理设备、面向多类型用户的电梯运行数据中台。融合百度飞浆深度学习框架下的AI模型快速开发、国产瑞芯rk3399 ARM系列的边缘处理技术应用、5G数据传输技术等关键技术,以技术创新解决现实生活中存在的需求。提出一个日常生活场景下综合应用人工智能技术、物联网技术和网络传输技术,解决现实存在的电梯运营中出现的困梯告警、消防告警、以至电动车进梯安全隐患。实现电梯运行的动态诊断和预测性维保,防患于未然。
Based on the huge maintenance demand generated by the extensive installation of elevator equipment,the elevator operation safety closely related to social life was taken as the scene,and comprehensively verified AI technology to realize elevator video data acquisition,rapid development of AI model,and AI system capability in domestic edge equipment.Research projects include:economic benefits and user experience of elevator operation data(sensors and video)transmission under 5G technology,development of elevator safety AI model(fire and electric vehicle detection),edge processing equipment based on domestic chips,and elevator operation data console for multiple types of users.Integrate key technologies such as the rapid development of AI model under the framework of Baidu flying plasma deep learning,the application of edge processing technology of domestic Ruixin rk3399 ARM series,5G data transmission technology,and solve the needs in real life with technological innovation.A comprehensive application of artificial intelligence technology,Internet of Things technology and network transmission technology in daily life scenarios were put forward to solve the practical hidden dangers of trapped elevator alarm,fire alarm and even electric vehicle entering elevator in elevator operation.The application realizes the dynamic diagnosis and predictive maintenance of elevator operation,and take preventive measures.
作者
马骁
何剑锋
尹中裕
Ma Xiao;He Jianfeng;Yin Zhongyu(China News Communication Construction Co.,Ltd.,Guangzhou 510030,China)
出处
《机电工程技术》
2022年第10期235-237,共3页
Mechanical & Electrical Engineering Technology
关键词
AI
5G
算法
边缘处理技术
智能管理平台
AI
5G
algorithm
edge processing technology
intelligent management platform
作者简介
第一作者:马骁(1979-),四川隆昌人,男,硕士,高级工程师(副教授级),研究领域为机电系统安装调试和建筑智能化技术,已发表论文4篇。
