A hybrid deep learning based traffic flow prediction method and its understanding

Deep neural networks (DNNs) have recently demonstrated the capability to predict traffic flow with big data. While existing DNN models can provide better performance than shallow models, it is still an open issue of making full use of spatial-temporal characteristics of the traffic flow to improve their performance. In addition, our understanding of them on traffic data remains limited. This paper proposes a DNN based traffic flow prediction model (DNN-BTF) to improve the prediction accuracy. The DNN-BTF model makes full use of weekly/daily periodicity and spatial-temporal characteristics of traffic flow. Inspired by recent work in machine learning, an attention based model was introduced that automatically learns to determine the importance of past traffic flow. The convolutional neural network was also used to mine the spatial features and the recurrent neural network to mine the temporal features of traffic flow. We also showed through visualization how DNN-BTF model understands traffic flow data and presents a challenge to conventional thinking about neural networks in the transportation field that neural networks is purely a “black-box” model. Data from open-access database PeMS was used to validate the proposed DNN-BTF model on a long-term horizon prediction task. Experimental results demonstrated that our method outperforms the state-of-the-art approaches.     

海员注意力研究

采用数字排列试验方法对348名海员的注意力进行测试，结果表明：(1)海员的注意力较好；(2)不同年龄、工龄和海龄组海员的注意力有显著差异；(3)不同职务、来源、文化程度的海员注意力有显著差异；(4)不同工种、婚，姻状况海员的注意力没有显著差异。     

公路卧铺客车卧铺结构浅析

对当前公路卧铺客车卧铺结构进行分析,并根据个人的经验提出在卧铺设计时应注意的事项。     

Detailed analysis of distraction induced by in-vehicle verbal interactions on visual search performance

We examined the negative effect of in-vehicle verbal interaction on visual search performance. Twenty participants performed a primary visual search task and a secondary verbal interaction task concurrently. We found that visual search performance deteriorated when the secondary task involving memory retrieval and speech production was performed concurrently. Moreover, a detailed analysis of the reaction time as a function of set size revealed that the increased reaction time was attributed not to the slowing of inspecting each item but to the increased processing time other than the inspection of each visual item, possibly due to task switching between the primary visual search task and the secondary verbal task. These findings have implications for providing information from in-vehicle information devices while reducing the risk of driver distraction.     

铁路机车司机驾驶能力的心理学检验指标及其效标

应用检验空军驾驶能力的心理学指标——五项纸笔方法对铁路机车司机驾驶能力进行心理学检验，并用标准９分制进行校标，结果表明除曲线轨迹相关较差不宜应用外，其他四项均可应用。     

基于站点实时关联度的短时公交客流预测方法

为探究公交站点之间的关联度并对公交客流进行更精准的实时预测，本文提出基于 Attention的交通预测核心算法(Traffic Forecast Model Based Attention，TFMA)，结合数据预处理和 站点信息编码完成基于站点实时关联度的短时公交客流预测方法。该方法首先创新性地提出了 站点实时关联度，可实现对目标站点客流量更精准的预测；其次，在公交站点的编码信息中融入 线路站点信息、客流变化率、天气、日期等关联因素；接着，该方法依靠Attention机制计算站点实 时关联度；核心算法中使用multi-headed机制、增加通道和残差连接进一步提升预测能力；最后， 以苏州市公交数据进行验证。结果显示：在准确率上，对比多元线性回归的53.8%、GRU(Gated Recurrent Unit)的66.9%和LightGBM(Light Gradient Boosting Machine)的81.2%，本文提出的基于 站点实时关联度的短时公交客流预测方法的准确率在90%以上，表明该方法具备优秀的短时公 交客流预测能力。     

沥青混凝土摊铺机的使用

介绍了沥青混凝土摊铺机在使用中应注意的问题及其相应的处理措施,以提高摊铺质量.