防爆电梯机房环境监测系统设计  

Design of Environment Monitoring System for Machine Rooms of Explosion-Proof Lifts

作　　者：倪敏敏 路成龙 Ni Minmin;Lu Chenglong

机构地区：[1]南京市特种设备安全监督检验研究院,江苏省南京市210002

出　　处：《工程机械》2025年第4期25-31,I0002,共8页Construction Machinery and Equipment

基　　金：江苏省市场监督管理局科技计划项目(KJ2023039)。

摘　　要：设计了一套机房环境监测系统,对机房温湿度、氧含量、气压及火焰烟雾进行监测。通过DHT11模块采集温湿度信息,通过SC03-O2模块采集含氧量信息,通过BMP180模块采集气压信息,采用K210模块采集视频信息,通过植入改进后的YOLOv8算法完成火焰、烟雾视觉识别。处理后的数据通过GPRS模块发送至远程监控平台。经验证,温湿度、大气压强和氧含量检测能够保证较高精度,同时改进后的算法mAP0.5达到了83.5%,mAP0.5∶0.95达到了51.7%,相对于原算法分别提高了3.1和3.2个百分点,Gflps和FPS指标也表现出色,证明了监控系统的有效性。A machine room environment monitoring system is designed to monitor the temperature,humidity,oxygen content,atmospheric pressure,flame and smoke in the machine room.Temperature and humidity information is collected through the DHT11 module,oxygen content information is collected through the SC03-O2 module,atmospheric pressure information is collected through the BMP180 module,video information is collected using the K210 module,and visual recognition of flame and smoke is accomplished by implanting the improved YOLOv8 algorithm.The processed data is sent to the remote monitoring platform through the GPRS module.It is verified that the temperature,humidity,atmospheric pressure and oxygen content detection can guarantee high precision,while mAP0.5 of the improved algorithm reaches 83.5%,and mAP0.5:0.95 reaches 51.7%,with a rise of 3.1%and 3.2%compared to the original algorithm,respectively,and the Gflps and FPS indicators also perform well,proving the effectiveness of the monitoring system.

关 键 词：防爆电梯 STM32 YOLOv8 视觉识别 算法改进 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]