基于改进 BiLSTM 网络的地铁车轮磨耗预测模型

针对地铁车轮磨耗数据时间跨度较长引起的长期依赖问题,为了进一步提升预测精度,提出一种将麻雀搜索算法(sparrow search algorithm,SSA)优化双向长短期记忆网络(bidirectional long short term memory,BiLSTM)的改进 BiLSTM(SSA-BiLSTM)网络模型,用于地铁车轮磨耗预测。首先,利用麻雀搜索算法对双向长短期记忆网络算法的神经元个数、迭代次数、输入批量和学习率等超参数在给定范围内进行寻优,得到参数最优值;然后,以参数最优值来构建改进 BiLSTM 网络模型,对车轮磨耗进行预测分析;最后,以车轮踏面磨耗和轮缘磨耗作为研究对象,将某地铁 1 车厢 1 号车轮的现场实测历史磨耗数据作为输入,对该模型进行训练及验证分析,并与多层感知机(multilayer perceptron,MLP)、LSTM、BiLSTM 以及 SSA-LSTM 模型的预测结果进行对比。研究结果表明：SSA-Bi-LSTM 模型的车轮磨耗预测精度更高,与 LSTM、BiLSTM 以及 SSA-LSTM 网络模型相比,踏面磨耗的平均绝对百分误差(mean absolute percentage error,MAPE)分别降低了 13.28%、10.32%、1.47%,轮缘磨耗分别降低了 9.5%、0.46%、0.02%;分别对同一地铁 2 号、4 号车厢的 1 号位置车轮磨耗进行预测,并与磨耗实测数据进行对比,踏面磨耗的平均绝对百分比误差分别为 1.34%、1.42%,轮缘磨耗的平均绝对百分比误差分别为 0.18%、0.19%,验证了本文所提模型具有良好的泛化性,为地铁轮对智能化管理提供理论支持,延长车轮使用寿命。     

Towards a better management of urban traffic pollution using a Pareto max flow approach

Rising levels of air pollution is a major concern across many parts of the world. In this article, we develop a transportation policy to handle air pollution caused by the heavy flow of traffic in urban areas. In particular, we aim to distribute the traffic flow more evenly through a city, by developing a flow algorithm that computes multiple solutions, each of which accommodates the maximum flow. The paper makes the following contributions to build such a transportation policy: (a) Develops a Pareto-optimal Max Flow Algorithm (PMFA) to suggest multiple max flow solutions. (b) Introduces the notion of k-optimality into PMFA to ensure that the suggested pareto solutions are sufficiently distinct from each other – referred to as Pareto-k-optimal Max Flow Algorithm (k-PMFA). (c) Through a series of experiments performed using the well-known traffic simulator SUMO and by doing emission modeling on the New York map, we could show that our policy distributes the air pollution more uniformly across locations.     

成都等交通物流融合发展经验与广州现代物流中心建设

本文根据实地调研成果,介绍了成都、重庆和郑州在建立健全物流管理体制机制、强化物流枢纽及通道功能、创新物流运输组织模式、物流标准化信息化、物流市场主体及产业培育等方面的先进经验,从加强政府顶层设计、完善物流枢纽和节点体系、促进物流枢纽互联互通、研究物流运输组织服务的新模式、打造“交通+物流+产业”的产业集群、提升物流标准化发展水平、出台促进物流降本增效的政策措施等七个方面总结了强化物流创新要素集聚,促进广州现代物流中心建设的经验和启示。     

A distributed VNS algorithm for optimizing dial-a-ride problems in large-scale scenarios

These days, transportation and logistic problems in large cities are demanding smarter transportation services that provide flexibility and adaptability. A possible solution to this arising problem is to compute the best routes for each new scenario. In this problem, known in the literature as the dial-a-ride problem, a number of passengers are transported between pickup and delivery locations trying to minimize the routing costs while respecting a set of prespecified constraints. This problem has been solved in the literature with several approaches from small to medium sized problems. However, few efforts have dealt with large scale problems very common in massive scenarios (big cities or highly-populated regions). In this study, a new distributed algorithm based on the partition of the requests space and the combination of the routes is presented and tested on a set of 24 different scenarios of a large-scale problem (up to 16,000 requests or 32,000 locations) in the city of San Francisco. The results show that, not only the distributed algorithm is able to solve large problem instances that the corresponding sequential algorithm is unable to solve in a reasonable time, but also to have an average improvement of 9% in the smaller problems. The results have been validated by means of statistical procedures proving that the distributed algorithm can be an effective way to solve high dimensional dial-a-ride problems.     

Quantifying the environmental impacts of increasing high occupancy vehicle lanes in the United States

To what extent will increasing High Occupancy Vehicle (HOV) lane-kilometer incentivize carpooling and reduce emissions of air pollutants and greenhouse gases? To answer these questions, we develop a multiple regression model relating HOV lanes and other socioeconomic factors to carpooling propensity in all 50 U.S. states and the District of Columbia, then calculate the extent to which increasing HOV lane-kilometers would lead to reductions in carbon dioxide equivalent (CO₂e) and major air pollutants across the U.S., by state. Increasing HOV lane extent has the greatest potential to reduce annual CO₂e in the District of Columbia, followed by Hawaii and New York. The smallest potential is found in states with the lowest population density, led by North Dakota. We then explore the extent to which recommendations made at one level of data aggregation (that of individual states) may be valid for another level, such as individual counties. The only state with sufficient data available to disaggregate the model to the county level is California, where we found a lower potential for state-wide CO₂e emission reductions under the county-level model as compared to the state-level model (0.69% as compared to 1.08%, under the same hypothetical scenario), albeit with significant differences in emission reduction potential between counties with higher vs. lower population densities. This analysis demonstrates the potential to generate generalizable insight into the magnitude of vehicle emission reductions that might be achieved through expanding HOV lanes, and highlights the importance of data disaggregation in identifying the optimal locations for potential reductions.     

Social Networks of Alaska Fishing Communities

Fishing community social networks function as channels for transfer of fishery knowledge, resources, and business transactions that help mitigate risks and shocks associated with altered access to fishery resources. Research on such networks in Alaska is limited despite their cultural importance and community reliance on fisheries. We contribute to scholarship of fishery social networks by assessing Alaska fishing community perspectives of challenges related to fisheries policy and management, and the existing social networks that aid in overcoming these challenges. Our findings show that the greatest challenges fishing communities face pertaining to fishery management are high costs of participating in catch share programs, restricted subsistence fishing activity due to decline in salmon, and complex regulations. Social networks exist for coping with these challenges; fishery information, and resources such as food, fuel and medicines are shared between communities. However, networks for accessing fishery support services are centralized in fewer larger communities and hubs such as Anchorage. Smaller and remote communities are the most compromised in this regard given the distance they must travel to access fishery support services. Leveraging social networks for sharing resources and improving fishery support services in smaller communities will increase their adaptive capacity and ability to maintain participation in Alaska fisheries.     

构建行业继续教育体系积极推进"人才强交"战略

坚持科学的人才观,牢固树立人才资源是第一资源,加强加快专业技术人员等交通从业人员继续教育和职业培训改革与发展,推进继续教育体系构建和培训模式创新,对于实施"科教兴交、人才强交"战略,实现交通新的跨越式发展具有重要的意义.     

基于改进PCNN模型的车辆阴影消除方法

消除车辆阴影对图像分割的影响是车辆自动监控系统中亟待解决的问题。针对阴影具有方向和大小不确定性的特点,充分利用脉冲耦合神经网络模型优良的图像分割能力,结合阴影属性,提出一种基于改进的脉冲耦合神经网络模型的运动车辆阴影消除方法。此方法无需建立道路背景模型与车辆阴影模型,在消除车辆阴影的同时能较好地保留车牌、车窗以及车标等车体的重要信息,为车体定位、特征提取等处理奠定良好的基础。大量现场图像测试结果验证了该方法消除车辆阴影的有效性,可广泛地应用于自然环境中车辆阴影的消除。     

太原铁路局危险货物运输安全专项整治对策

阐述太原铁路局危险货物运输现状,分析当前危险货物运输安全管理中存在的主要问题,提出加强危险货物运输安全的措施：全部实现源头＂机控＂把关;危险货物运输清理工作应全部到位;全面推进安全现状综合分析和安全评价工作;提升危险货物运输风险控制管理水平;提高应急救援的装备和应对能力;完善危险货物运输培训覆盖面,提高培训水平。