摘要
安全生产管理是建筑施工中重要的组成,为了降低施工过程中导致的风险,提升建筑施工管理中信息化的水平,本文首先设计了基于人工智能的施工安全生产监控与决策管理系统,并详细分析了系统的各个软硬件设计方案和构成。随后,本文针对施工现场中烟火等目标检测中存在的检测模型庞大,检测效果不佳的问题,提出一个基于轻量级神经网络ShuffleNet的施工环境下烟火检测模型。该模型检测精度达到95.7%,能对各个环境中的烟火区域进行十分高效的感知。最后,本文对提出的模型进行了定量和定性的可视化分析,以验证模型的感知效果。实验表明,本文提出的模型能够有效的定位施工环境中出现的烟火区域。
Safety production management is an important component in construction.In order to reduce the risks caused by the construction process and to improve the level of information technology in construction management,this paper first designs a con-struction safety production monitoring and decision management system based on artificial intelligence,and analyses in detail the indi-vidual hardware and software design solutions and composition of the system.Subsequently,this paper proposes a smoke and fire de-tection model based on ShuffleNet,a lightweight neural network,to address the problems of large detection models and poor detection effects in the detection of targets such as smoke and fire in construction sites.The model achieves a detection accuracy of 95.7%and can sense the smoke and fire areas in each environment very efficiently.Finally,a quantitative and qualitative visualisation analysis of the proposed model is carried out to verify the perceptual effectiveness of the model.The experiments show that the proposed model can effectively locate the smoke and fire areas in the construction environment.
作者
曾强
冯身强
ZENG Qiang;FENG Shengqiang(Chengdu Xingcheng Investment Group Co.,Ltd.,Chengdu 610000,China;Chengdu Construction Industrial Construction Co.,Ltd.,Chengdu 610000,China)
出处
《自动化与仪器仪表》
2023年第6期156-160,共5页
Automation & Instrumentation
关键词
安全生产
智慧施工
管理系统
safe manufacturing
smart construction
management systems