摘要
随着物联网技术发展,公路工程质量检测及施工工艺已实现数据自动采集及传输。本文通过关联分析的方法,对沥青路面施工中施工工艺与检测技术进行了研究,分析不同数据建模原理,并采用CNN-Bi LSTM多任务学习模型进行相关性研究。依托京沪高速公路新沂至淮安段改扩建工程,采集7 500条数据作为样本,选取1 000条数据作为模型测试集。结果表明,CNN-Bi LSTM模型精准率最高达到了0.944,召回率最高达到了0.897,由此得出沥青路面压实度与施工工艺所选参数相关性很高,可用于预测沥青路面压实度。
With the development of Internet of Things technology,highway engineering quality inspection and construction technology have achieved automatic data collection and transmission. Through the method of correlation analysis,this paper studies the construction technology and detection technology in asphalt pavement construction,analyzes different data modeling principles,and uses the CNN-BiLSTM multi-task learning model to conduct correlation research. Relying on the reconstruction and expansion project of the Xinyi-Huai’an section of the Beijing-Shanghai Expressway,7 500 pieces of data were collected as samples,and 1 000 pieces of data were selected as the model test set. The results show that the accuracy rate of the CNN-BiLSTM model is up to 0.944,and the recall rate is up to 0.897. It is concluded that the compaction degree of the asphalt pavement is highly correlated with the parameters selected for the construction process,and the compaction of the asphalt pavement can be predicted in advance.
作者
朱玉
ZHU Yu(Anhui Transportation Holding Group Co.,Ltd.,Hefei Anhui 230000,China)
出处
《交通节能与环保》
2022年第5期155-159,共5页
Transport Energy Conservation & Environmental Protection
