基于机器视觉的车辆检测与参数识别研究进展

近年来公路交通运输快速增长,交通车辆的快速准确检测与识别对智能交通系统和交通基础设施运维具有重要意义.随着机器视觉和深度学习技术的迅速发展及其在目标检测领域的广泛应用,车辆目标检测和参数识别也取得新的突破.该文从车辆参数的识别方法和应用研究两方面梳理了机器视觉和深度学习在车辆检测与参数识别领域的研究现状、最新研究成果和...     

基于Faster R-CNN与形态法的路面病害识别

为提高基于图像处理的路面表观病害检测识别效率及精度,引入目标检测中的快速区域卷积神经网络(Faster Region Convolutional Neural Network,Faster R-CNN)算法以快速识别病害种类、位置与面积;针对已提取的带边框裂缝病害区域,采用基于VGG16迁移学习与模型微调的CNN与50...     

在役轨道交通矮塔斜拉桥检测与评估

以某在役轨道交通矮塔斜拉桥的检测与评估为例,通过对矮塔斜拉桥受力特点的分析,确定了定期检测中矮塔斜拉桥的检测重点.主梁、主塔和墩柱的恒载几何形态测量表明,该矮塔斜拉桥整体保持稳定的状态,主梁跨中有沉降的趋势;通过桥梁自振特性、斜拉索索力和动位移的测试,明确了该桥的动力特性和动力响应;通过外观检查,明确了主梁主塔混凝土结...     

公路隧道照明精确检测与智能控制系统研究

为解决公路隧道照明系统节能与行车安全、运营管理之间的矛盾,采用理论分析及实际检验的方法,分析了隧道照明控制系统的调光原理、照明控制系统结构以及亮度检测误差产生的原因,设计了一种可调量程的亮度检测仪并给出校准及亮度检测步骤,解决了公路隧道亮度精确检测的难题,采用模糊神经网络算法在照明系统中进行了实际应用,实践证明本算法在亮度满足实际运营需求的前提下可取得良好的节能效果。     

A signal-based fault detection and classification method for heavy haul wagons

This paper proposes a signal-based fault detection and isolation (FDI) system for heavy haul wagons considering the special requirements of low cost and robustness. The sensor network of the proposed system consists of just two accelerometers mounted on the front left and rear right of the carbody. Seven fault indicators (FIs) are proposed based on the cross-correlation analyses of the sensor-collected acceleration signals. Bolster spring fault conditions are focused on in this paper, including two different levels (small faults and moderate faults) and two locations (faults in the left and right bolster springs of the first bogie). A fully detailed dynamic model of a typical 40t axle load heavy haul wagon is developed to evaluate the deterioration of dynamic behaviour under proposed fault conditions and demonstrate the detectability of the proposed FDI method. Even though the fault conditions considered in this paper did not deteriorate the wagon dynamic behaviour dramatically, the proposed FIs show great sensitivity to the bolster spring faults. The most effective and efficient FIs are chosen for fault detection and classification. Analysis results indicate that it is possible to detect changes in bolster stiffness of ±25% and identify the fault location.     

Shockwave-based queue estimation approach for undersaturated and oversaturated signalized intersections using multi-source detection data

With the progress of information and sensing technologies, estimating vehicular queue length at signalized intersections becomes feasible and has attracted considerable attention. The existing studies provided a solid theoretical foundation for the estimation; however, the studies have some restrictions or limitations more or less. This paper presents a new methodology for estimating vehicular queue length at signalized intersections using multi-source detection data under both undersaturated and oversaturated conditions. The methodology applies the shockwave theory to model queue dynamics. Using data from probe vehicles and point detectors, analytical formulations for calculating the maximum and minimum (residual) queue lengths of each cycle are developed. Ground truth data were collected from numerical experiments conducted at two intersections in Shanghai, China, to verify the proposed methodology. It is found that the methodology has mean absolute percentage errors of 17.09% and 12.28%, respectively, for maximum queue length estimation in two tests, which are reasonably effective. However, the methodology is unsatisfactory in estimating the residual queue length. Other limitations of the proposed models and algorithms are also discussed in the paper.     

Plate scanning tools to obtain travel times in traffic networks

The management of vehicle travel times has been shown to be fundamental to traffic network analysis. To collect travel time measurement, some methods focus solely on isolated links or highway segments, and where two measurement points, at the beginning and at the end of a section, are deemed sufficient to evaluate users' travel time. However, in many cases, transport studies involve networks in which the problem is more complex. This article takes advantage of the plate scanning technique to propose an algorithm that minimizes the required number of registering devices and their location in order to identify vehicles candidates to compute the travel times of a given set of routes (or subroutes). The merits of the proposed method are explained using simple examples and are illustrated by its application to the real network of Ciudad Real.     

基于Kinect深度虚拟线圈的夜间车流量检测

针对夜间环境下基于摄像机的车辆检测方法存在精度低、稳定性差以及无法对车型进行有效识别等问题,提出一种基于Kinect深度虚拟线圈的夜间车辆检测与计数算法.首先对Kinect深度图像进行预处理,分别获得运动目标深度图(MDM)与空洞深度图(HDM).然后在MDM与HDM上设置虚拟线圈,利用积分图像分别生成对应的一维运动信号,对其进行加权合成获得对车辆运动特征的表达,并在合成的运动信号范围内检测出车辆目标,并计算出车辆目标的几何特征,通过SVM对车型进行有效识别.实验结果表明,该算法对于单双车道的车辆计数正确率分别高达99.75%与99.25%,大小车型分类正确率可达99.80%,处理单张图片的平均时间仅为7 ms.     

Bolster spring fault detection strategy for heavy haul wagons

An on-board health monitoring system is proposed for heavy haul wagons in this paper including a signal-based fault detection and isolation (FDI) method and an on-line fault diagnose strategy. Such a system, to be feasible on freight wagons, must be sufficiently cheap and robust, hence the design assumes the constraint of using only two accelerometers mounted on the front left and right rear part of each carbody in a heavy haul train. This paper focuses on the detection of bolster spring fault conditions. The problem is made more complex by the modes of failure which might be expected in bolster spring nests. Types of spring failure are firstly identified and discussed covering situations of broken (shortening springs) and softening (individual spring loss from a nest or cross-section loss through corrosion). The effects of these faults and their detectability were investigated using simulations on straight and curved track and using a fully detailed model of a typical 40?t axle-load heavy haul wagon. The simulation results were then examined and compared using cross-correlation analysis and an FDI system was proposed. The FDI system introduced five possible fault indicators. Initial results indicated that it was possible to detect changes in bolster stiffness of ±25%. An on-line fault diagnoses strategy is proposed for bolster spring faults which only requires data from wagon monitoring during travel around sharp curves to detect and the occurrence of confirm faults. The functionality envisaged needs only a ‘once per trip’ monitoring site, such as a sharper curve, and is aimed at condition monitoring rather than the provision of alarms or comprehensive monitoring of all events.     

汽车离合器从动盘扭转耐久试验中弹簧疲劳断裂的自动识别

对汽车离合器从动盘扭转耐久试验中弹簧断裂的自动识别技术进行研究，提出一种以弹簧刚度特性变化为识别特征的识别原理，进行了相关的分析计算；介绍了基于此原理的自动识别方案及测试过程，并给出实际运行结果。此技术提高了测试智能化水平。