铁路车辆修理布局集中化研究

针对全路车辆修理量逐渐减少,各修理点修理能力严重过剩的现状,提出铁路车辆修理布局集中化方法。分析铁路车辆修理布局集中化的内涵,提出铁路车辆修理布局集中化双层规划模型的构建思路;以修理点改扩建费用、运营费用、运输费用之和最小为目标构建上层规划模型,以车辆运输费用最小为目标构建下层规划模型,并利用迭代思想,设计基于改进后的蚁群算法的双层迭代优化算法求解该双层规划模型。借助MATLAB软件进行算例分析,验证了模型的有效性。     

增强边缘梯度特征局部量化策略驱动下的车标识别

目的 车标是车辆的显著性特征,通过车标的分类与识别可以极大缩小车辆型号识别的范围,是车辆品牌和型号识别中的重要环节。基于特征描述子的车标识别算法存在如下缺点：一方面,算法提取的特征数量有限,不能全面描述车标的特征;另一方面,提取的特征过于冗杂,维度高,需要大量的计算时间。为了提取更加丰富的车标特征,提高识别效率,提出一种增强边缘梯度特征局部量化策略驱动下的车标识别方法。方法 首先提取车标图像的增强边缘特征,即根据不同的梯度方向提取梯度信息,生成梯度大小矩阵,并采用LTP（local ternary patterns）算子在梯度大小矩阵上进一步进行特征提取,然后采用特征码本对提取的特征进行量化操作,在确保车标特征描述能力的同时,精简了特征数目,缩短了局部向量的长度,最后采用WPCA（whitened principal component analysis）进行特征降维操作,并基于CRC（collaborative representation based classification）分类器进行车标的识别。结果 基于本文算法提取的车标特征向量,能够很好地描述车标图像的特征,在HFUT-VL1车标数据集上取得了97.85%的识别率（平均每类训练样本为10张）,且在识别难度较大的XMU车标数据集上也能取得90%以上的识别率（平均每类训练样本为100张）,与其他识别算法相比,识别率有明显提高,且具有更强的鲁棒性。结论 增强边缘梯度特征局部量化策略驱动下的车标识别算法提取的特征信息能够有效地描述车标,具有很高的识别率和很强的鲁棒性,大大降低了特征向量的维度,提高了识别效率。     

特种车辆维修空间可达性评价方法研究

可达性作为维修性的定性指标,缺少定量评价的方法。传统方法只针对装备的整体进行定性评价,不能对不同的等级维修和单个维修动作进行评价。通过对维修性设计中的结构性因素进行分析,得到了空间可达性的主要影响因素。在虚拟维修仿真平台的基础上对工具旋转角度和作业空间比进行了量化评分,给出了针对维修作业过程中单个零部件拆卸的空间可达性评价方法。单个动作评价方法为等级维修的整体空间可达性评价提供了数据基础,并为新型特种车辆的维修方案论证和研制的空间可达性评价提供了手段。     

基于贝叶斯网络的两栖车辆海上航行安全评估研究

分析了两栖车辆海上航行相关风险,进行安全风险评估,并对可能发生的风险提出了一种可行的解决方案。基于贝叶斯点估计和贝叶斯学习估计与两栖车辆海上航行有关的交通事故,建立了两栖车辆海上航行交通系统(QRA)分析模型。通过贝叶斯网络来获取其相对风险的QRA。研究结果显示：使用该方法描述了两栖车辆海上航行的风险分布,并在QRA上显示了与各种特征相关结果,对两栖车辆航行构成相对风险综合分析,高风险区域的规模按降序排列如下：1)航道—引航—国内水道—航行;2)沿海地区—非引航区—国家水道—航行。该评估模型具有较好的应用价值,可为两栖车辆海上安全航行提供借鉴。     

抗模糊特征提取策略下的车标识别

目的 现有的车标识别方法尽管取得了不错的识别效果,但最终的识别率容易遇到瓶颈,很难得到提升。车标识别是智能交通系统中至关重要的一部分,识别率的微小提升也能带来巨大的社会价值。通过挖掘与分析车标识别中潜在的问题和难点,发现未能得到正确分类的图像大部分为模糊车标图像。针对车标图像中存在的成像模糊等情况,本文提出一种基于抗模糊特征提取的车标识别方法。方法 构建车标图像金字塔模型,分别提取图像的抗纹理模糊特征和抗边缘模糊特征。抗纹理模糊特征的提取使用局部量化的LPQ（local phase quantization）模式,可以增强原始特征的鲁棒性,抗边缘模糊特征的提取基于局部块弱梯度消除的HOG（histogram of oriented gradient）特征提取方法,可以在描述车标图像边缘梯度信息的同时,提升特征的抗模糊能力。最后利用CCA（canonical correlation analysis）方法进行两种抗模糊特征的融合并用于后续的降维与分类。结果 本文方法在多个数据集上均取得了很好的识别效果,在20幅训练样本下,本文方法在公开车标数据集HFUT-VL（vehicle logo dataset from Hefei University of Technology）上取得了99.04%的识别率,在本文构建的模糊车标数据集BVL（blurring vehicle logo dataset）上也取得了97.19%的识别率。而在难度较大的XMU（Xiamen University vehicle logo dataset）上,本文方法在100幅训练样本下也达到了96.87%的识别率,识别效果高于一些具有较好表现的车标识别方法,表现出很强的鲁棒性和抗模糊性。结论 本文方法提高了对成像质量欠缺的车标图像的识别能力,从而提升了整体识别效果,更符合实际应用中车标识别的需求。     

基于改进K—Means算法的车辆识别方法

车辆的检测和识别一直是道路监控、安全辅助驾驶、车辆自主导航等领域的重要研究内容。文中基于机器视觉方法,在Lab颜色空间通过对经典K—Means算法的聚类中心、聚类数和距离测度三方面的改进实现了L分量的聚类,从而达到图像分割的目的。提取图像的矩形度、Hu矩和Affine矩的特征,针对不同的车型建立各自的模板,利用改进Hu不变矩和仿射不变矩的组合不变矩对分割后图像进行车型的识别。实验结果表明,文中提出的方法对于复杂环境下的车辆检测和识别具有良好的可靠性和鲁棒性。     

A fuzzy approach for the reviewer assignment problem

Assignment of experts to project proposals is a significant task for funding agencies which have to assess the potential value of the research and development (R&D) projects through peer review. The problem is known as reviewer assignment problem and has real-world applications in funding agencies, conferences and research journals. Given a set of experts and a set of proposals; the problem can be defined as assigning the most suitable experts to the proposals under some constraints, which are generally encountered by funding agencies. In this study, a fuzzy model is offered to solve the reviewer assignment problem. The objective of the model is to maximize the total matching degree of assigned experts under some constraints such as cost of forming a panel and the size of a panel. The matching degrees are defined using linguistic variables to denote the expertise of each expert with respect to each proposal. The fuzzy mathematical model, which also takes into account different constraints related to the problem, is solved via the selected fuzzy ranking methods namely; the signed distance method and the method of ranking fuzzy numbers with integral value. The solution of an example problem – inspired from a real-life situation – with both of the mentioned methods revealed the effectiveness of the solution approach. It is believed that the use of the offered fuzzy approach could improve the accuracy of the decisions made by funding agencies.     

Effects of agricultural quad bike driving on postural control during static,dynamic and functional tasks – A field study

BackgroundDriving a quad bike in a rural occupational setting is likely to expose the driver to various physical stimuli including whole-body vibration (WBV). These exposures may be linked to post-driving postural alterations which in turn could lead to an increased risk of spinal injury while undertaking manual material handling activities immediately following driving or falls while exiting from a vehicle.PurposeThe purpose of this study was to use a battery of postural tasks namely; bipedal and unipedal stance, limits of stability (LOS) and lifting task to assess how quad bike driving alters the postural control (PC) in a group of rural workers.MethodsThe PC, determined from centre of pressure (COP) displacements in a group of rural workers (n = 34), was evaluated at three time (T) periods, once (T-I) before the 30 min quad bike driving session on a typical New Zealand farm terrain and twice (T-II and T-III) immediately following driving, each time period lasted approximately 10 min later.ResultsThe results demonstrated a significant (p < 0.05) increase and decrease in the magnitude of the COP measures for the lifting task and unipedal stance respectively during both T-II and T-III periods. However significant (p < 0.05) increase in the magnitude COP measures of bipedal stance, and increase in the maximal stability limits (LOS) were demonstrated only in the T-III.ConclusionsThese results demonstrate both immediate and sustained alterations in the PC following a period of occupational vehicle driving. Also, the results demonstrate both a worsening and improvement in postural control during the performance of a battery of tasks. Observed adverse or facilitatory postural effects will require further laboratory based investigations in order to determine how such disparity can best be explained or challenged.Relevance to industryThe findings will inform ergonomists about the potential risk involved in carrying out physically demanding occupational tasks following vehicle driving. This putative situation could be prevented by adopting behavioural strategies by drivers, and engineered interventions designed to reduce WBV exposure.     

The influence of company identity on the perception of vehicle sounds

In order to determine how the interior of a car should sound, automotive manufacturers often rely on obtaining data from individual evaluations of vehicle sounds. Company identity could play a role in these appraisals, particularly when individuals are comparing cars from opposite ends of the performance spectrum. This research addressed the question: does company identity influence the evaluation of automotive sounds belonging to cars of a similar performance level and from the same market segment? Participants listened to car sounds from two competing manufacturers, together with control sounds. Before listening to each sound, participants were presented with the correct company identity for that sound, the incorrect identity or were given no information about the identity of the sound. The results showed that company identity did not influence appraisals of high performance cars belonging to different manufacturers. These results have positive implications for methodologies employed to capture the perceptions of individuals.

Statement of Relevance: A challenge in automotive design is to set appropriate targets for vehicle sounds, relying on understanding subjective reactions of individuals to such sounds. This paper assesses the role of company identity in influencing these subjective reactions and will guide sound evaluation studies, in which the manufacturer is often apparent.     

多任务分段紧凑特征的车辆检索方法

目的 随着公共安全领域中大规模图像监控及视频数据的增长以及智能交通的发展,车辆检索有着极其重要的应用价值。针对已有车辆检索中自动化和智能化水平低、难以获取精确的检索结果等问题,提出一种多任务分段紧凑特征的车辆检索方法,有效利用车辆基本信息的多样性和关联性实现实时检索。方法 首先,利用相关任务之间的联系提高检索精度和细化图像特征,因此构造了一种多任务深度卷积网络分段学习车辆不同属性的哈希码,将图像语义和图像表示相结合,并采用最小化图像编码使学习到的车辆的不同属性特征更具有鲁棒性;然后,选用特征金字塔网络提取车辆图像的实例特征并利用局部敏感哈希再排序方法对提取到的特征进行检索;最后,针对无法获取查询车辆目标图像的特殊情况,采用跨模态辅助检索方法进行检索。结果 提出的检索方法在3个公开数据集上均优于目前主流的检索方法,其中在CompCars数据集上检索精度达到0.966,在VehicleID数据集上检索精度提升至0.862。结论 本文提出的多任务分段紧凑特征的车辆检索方法既能得到最小化图像编码及图像实例特征,还可在无法获取目标检索图像信息时进行跨模态检索,通过实验对比验证了方法的有效性。     

基于AMR传感器的行驶车辆检测分类算法*

深入研究了各向异性磁阻（AMR）传感器的数据采集原理及特征波形向量提取方法,提出基于AMR传感器及加权欧氏距离的车辆分类识别算法。道路车辆检测实验数据显示,与感应线圈车辆检测法及视频车辆检测等方法相比,该检测方法基本不受环境路况天气等外在因素的影响,满足长期稳定精确等车辆检测的要求,同时在性能、成本、寿命、实时性、维护和升级等方面整体上有较大的优越性。     

基于无线网络的车辆健康状态管理系统研究

针对特种车辆领域售后服务过程高效率、高可靠性、体系化的迫切需求,分析了当前国内外汽车领域特别是特种车辆领域售后服务过程广泛存在的问题,通过研究无线智能终端、车联网技术、PHM等关键技术,重点论述了基于无线网络的车辆健康状态管理系统构建的意义及构建策略,以及如何将该系统应用于车辆售后服务工作,并阐述了该系统的初步设计方案;最后经过车辆的实际试验,证明该系统符合工程机械车辆售后维护工作要求,所建设的故障诊断和健康评估模型具有实用化推广价值。     

Ergonomics risk assessment among maintenance operators in a Tunisian railway company: A case study

Railway truck maintenance plays a crucial role for rail transport. It aims at ensuring the availability of the equipment and the safety of the operations. However, maintenance activities often involve uncommon work and nonroutine tasks and it is obvious that maintenance operators are more threatened than other employees to be exposed to various hazards in their workplace. In Tunisia, the prevalence of occupational accidents and injuries among maintenance workers is not well documented. The target of this cross‐sectional study is to assess ergonomics risks associated with maintenance activities in a railway company. Data collection was carried out through a self‐administrated questionnaire and via direct observations. The Body Parts Symptom Survey (BPSS) was used to identify the discomfort level of the workers. More than 100 maintenance operators working in nine different workshops were involved in this study. From BPSS results, this paper proposed a new method to calculate the discomfort index related to each body part in order to calculate the global discomfort index (GDI) for each workshop. Results showed that the most hazardous workshop was rail‐car maintenance workshop (GDI = 43.06). Four main tasks were assessed with normalized score. Final scores were found to be extremely high. To remedy the raised problems, new adjustable workstations were proposed, which showed a considerable improvement on workers’ postures.     

Advanced Interfaces for Vehicle Teleoperation: Collaborative Control, Sensor Fusion Displays, and Remote Driving Tools

We are working to make vehicle teleoperation accessible to all users, novices and experts alike. In our research, we are developing a new control model for teleoperation, sensor-fusion displays and a suite of remote driving tools. Our goal is to build a framework which enables humans and robots to communicate, to exchange ideas and to resolve differences. In short, to develop systems in which humans and robots work together and jointly solve problems.     

虚拟视景中的装甲车辆运动仿真算法研究

通过分析履带式装甲车的行动装置与运动特点,研究了装甲车在任意地形上的运动情况,以满足分布交互式虚拟仿真实时性和逼真度要求对实体行为进行了建模与简化,选取适合于履带式车辆的支撑点来研究车辆的姿态问题,提出了在任意地形中的运动仿真算法来确定运动后的位置和姿态.在基于VTP平台开发的装甲对抗仿真系统中,该算法实现了装甲车的运动仿真,并能很好的满足系统实时性和高逼真度的需求.     

基于Fast R-CNN的车辆目标检测

目的 在传统车辆目标检测问题中,需要针对不同图像场景选择适合的特征。为此提出一种基于快速区域卷积神经网络（Fast R-CNN）的场景图像车辆目标发现方法,避免传统车辆目标检测问题中需要设计手工特征的问题。方法 该方法基于深度学习卷积神经网络思想。首先使用待检测车辆图像定义视觉任务。利用选择性搜索算法获得样本图像的候选区域,将候选区域坐标与视觉任务示例图像一起输入网络学习。示例图像经过深度卷积神经网络中的卷积层,池化层计算,最终得到深度卷积特征。在输入时没有规定示例图像的规格,此时得到的卷积特征规格不定。然后,基于Fast R-CNN网络结构,通过感兴趣区域池化层规格化特征,最后将特征输入不同的全连接分支,并行回归计算特征分类,以及检测框坐标值。经过多次迭代训练,最后得到与指定视觉任务强相关的目标检测模型,具有训练好的权重参数。在新的场景图像中,可以通过该目标检测模型检测给定类型的车辆目标。结果 首先确定视觉任务包含公交车,小汽车两类,背景场景是城市道路。利用与视觉任务强相关的测试样本集对目标检测模型进行测试,实验表明,当测试样本场景与视觉任务相关度越高,且样本中车辆目标的形变越小,得到的车辆目标检测模型对车辆目标检测具有良好的检测效果。结论 本文提出的车辆目标检测方法,利用卷积神经网络提取卷积特征代替传统手工特征提取过程,通过Fast R-CNN对由示例图像组成定义的视觉任务训练得到了效果良好的车辆目标检测模型。该模型可以对与视觉任务强相关新场景图像进行效果良好的车辆目标检测。本文结合深度学习卷积神经网络思想,利用卷积特征替代传统手工特征,避免了传统检测问题中特征选择问题。深层卷积特征具有更好的表达能力。基于Fast R-CNN网络,最终通过多次迭代训练得到车辆检测模型。该检测模型对本文规定的视觉任务有良好的检测效果。本文为解决车辆目标检测问题提供了更加泛化和简洁的解决思路。     

纯电动车车身控制CANopen协议设计

针对纯电动汽车车身控制系统,提出其应用层CANopen协议的设计方案。分析车身控制系统中各电子控制单元的功能。根据 CANopen协议要求,设计各电子控制单元的网络管理报文对象、过程数据对象、服务数据对象、特殊功能对象。通过CANoe软件及其组件CANoe.CANopen建立纯电动汽车车身控制CANopen网络仿真平台,采用CAPL语言编写程序模拟真实节点功能,运用Panel Editor软件编辑控制面板。仿真结果表明,CANopen协议可以满足纯电动汽车车身控制系统的控制需求。     

Experiments in the coordinated control of an underwater arm/vehicle system

The addition of manipulators to small autonomous underwater vehicles (AUVs) can pose significant control challenges due to hydrodynamic interactions between the arm and the vehicle. Experiments conducted at the Monterey Bay Aquarium Research Institute (MBARI) using the OTTER vehicle have shown that dynamical interactions between an arm and a vehicle can be very significant. For the experiments reported in this paper, a single-link arm was mounted on OTTER. Tests showed that for 90-degree, two-second repetitive slews of the arm, the vehicle would move as much as 18 degrees in roll and 14 degrees in yaw when no vehicle control was applied.Using a new, highly accurate model of the arm/vehicle hydrodynamic interaction forces, which was developed as part of this research, a coordinated arm/vehicle control strategy was implemented. Under this model-based approach, interaction forces acting on the vehicle due to arm motion were predicted and fed into the vehicle controller. Using this method, station-keeping capability was greatly enhanced. Errors at the manipulator end point were reduced by over a factor of six when compared to results when no control was applied to the vehicle and by a factor of 2.5 when compared to results from a standard independent arm and vehicle feedback control approach. Using the coordinated-control strategy, arm end-point settling times were reduced by a factor three when compared to those obtained with arm and vehicle feedback control alone. These dramatic performance improvements were obtained with only a five-percent increase in total applied thrust.     

Hybrid-augmented intelligence in predictive maintenance with digital intelligent assistants

Industrial maintenance strategies increasingly rely on artificial intelligence to predict asset conditions and prescribe maintenance actions. The related maintenance software and human maintenance actors can form a hybrid-augmented intelligence system where each side benefits from and enhances the other side's intelligence. This system requires optimized human-machine interfaces to help users express their knowledge and retrieve information from difficult-to-use software. Therefore, this article proposes a novel approach for maintenance experts and operators to interact with a predictive maintenance system through a digital intelligent assistant. This assistant is artificial intelligence (AI) that could help its users interact with the system via natural language and collect their feedback about the success of maintenance interventions. Implementing hybrid-augmented intelligence in a predictive maintenance system faces several technical, social, economic, organizational, and legal challenges. The benefits, limitations, and risks of hybrid-augmented intelligence must be clear to all employees to advocate its use. AI-focused change management and employee training could be techniques to address these challenges. The success of the proposed approach also relies on the continuous improvement of natural language understanding. Such a process will need conversation-driven development where actual interactions with the assistant provide accurate training data for language and dialog models. Future research has to be interdisciplinary and may cover the integration of explainable AI, suitable AI laws, operationalized trustworthy AI, efficient design for human-computer interaction, and natural language processing adapted to predictive maintenance.     

An analysis of experts' ratings of ergonomic stress

Experts in the field of industrial ergonomics are frequently called upon to visit workplaces in order to evaluate and rate ergonomic stressors which may cause fatigue, discomfort, or injury. The objective of this study was to evaluate the extent of agreement among experts' perceptions of four categories of ergonomic stress: materials handling (e.g., lifting), upper extremity activities, posture, and energy expenditure.

A panel of five experts observed ten jobs in manufacturing and warehouse facilities. A three-point scale (1 = insignificant stress, 2 = moderate stress, 3 = high stress) was used to rate the jobs in each of the four stress categories. In addition, lifting stresses on each job were analyzed using methods described in the NIOSH Work Practices Guide for Manual Lifting. The resulting data were analyzed to determine the level of expert consensus and the level of agreement with the NIOSH results.

In general, there was a high level of consensus among the experts. Perfect agreement (identical ratings by all five experts) was obtained for 30 percent of the experts' scores; while consensus agreement (identical ratings by three or more experts) was obtained for 87.5 percent of the scores. In general, the experts' perceptions agreed with the NIOSH ratings of lifting stress. On a few jobs, however, the expert panel considered lifting stresses to be greater than predicted by the NIOSH ratings.