高速公路养护维修作业区是高速公路的瓶颈路段， 也往往成为交通事故的黑点地段， 其高速的交通流和复杂的现场作业环境使运行车辆和作业人员具有高风险性。作业区不仅给高速公路交通系统带来不利影响，也给驾驶人员和作业人员带来安全隐患。 如何改善高速公路养护维修作业区的交通安全状态一直是交通学者和高速公路管理部门的研究重点， 而提出合理的作业区交通组织方案和管理对策的前提是研究高速公路养护维修作业区的车辆运行特性。 本文利用无人机航拍高速公路养护维修作业区的交通流视频，提取相关数据，并考虑车道和作业区上游到工作区的不同距离断面等因素，研究作业区的车辆运行特性。研究方法和研究结论如下：
       一、 从微观角度分析作业区合流段内不同断面各车道上交通量分担率和车速的变化规律，与从宏观角度分析作业区事故数据互补，宏观态势研究和微观机理分析相结合发现距离上游过渡区起点200m到工作区起点之间的车辆运行特性有显著改变。 此区域内不同断面各车道上的交通量分担率变化明显、 平均车速降低，车辆运行的风险性较大，为高速公路养护维修作业区的行车风险区域。
       二、 建立自由流状态下作业区行车风险区域内的车速预测模型与研究不同断面各车道上的速度-流量关系形成互补，分析作业区行车风险区域内不同断面、不同车道、不同流量下的车速特性和相邻断面间的车速变化规律。研究发现作业区对车速特性的影响在距离上游过渡区 30m 到缓冲区起点的区域内最为明显。
       三、 通过数据分析和建模研究了作业区行车风险区域内不同断面各车道的交通量分布特性、车辆的运行轨迹、换道起始点、换道结束点、以及车辆换道长度等内容。 研究发现作业区行车风险区域内的车辆换道行为存在一种横向延迟扩散现象，换道行为先从封闭车道开始，逐渐向相邻车道扩散。另外，随着换道起始点靠近上游过渡区起点，车辆的换道长度呈下降趋势。

    Work zones are the bottleneck road of freeway, often also become the black-spot sections, this is because that the fast traffic volume and complex operating environment make vehicle running and workers have a high risk. Work zones not only have negative effects on the whole freeway transportation system, but also bring hidden danger to drivers and workers. How to improve the safety level of work zone has been the research focus of traffic scholars and management department. However, the premise of putting forward optimal traffic organization scheme and management on work zones is a good understanding of the driving characteristics of vehicles. In this paper, a representative work zone is shot by using an unmanned aerial vehicle (UAV), then relevant data is extracted, and vehicle driving characteristics are studied by taking the deviations of lane and section with different distance to the work zone into considerarion. The research methods and conclusions are as follows.
    First, from a micro-scale analysis of the traffic distribution and the speed characteristic of vehicles among lanes on each section of work zone, which is combined with a macro-scale analysis of the accident data. It is found that there is a significiant change in the vehicle operating characteristics from 200 meters distant from the upstream transition zone to work zone. In this area, the traffic sharing ratio of different sections in each lane changes obviously, the average speed reduces, the risk of vehicle operation is relatively large. Therefore, this area is the driving risk area of work zone.
    Second, establishment of speed prediction model in the driving risk area under free flow is combined with study of the velocity and flow relation of different section in different lanes, this paper analyses the change of speed in defferent sections, different lanes and different traffic conditions in the driving risk area of work zone. The most obvious influence of work zone on the speed characteristics is from 30 meters distant from the upstream transition zone to the buffer zone.
    Third, the change of traffic sharing ratio of different sections in different lanes, the vehicle moving trajectory, starting point of lane chaning, end point of lane changing and lane changing length are studied based on data analysis and modeling. There is a phenomenon of horizontal delay diffusion in the vehicle lane changing behavior of each lane in the driving risk area. That is lane changing occurs first on the closed lane, gradually spread to the adjacent lane. As the starting point of lane changing is closed to the upstream transition zone, the length of the lane changing is decreased.